The item \"%s\" has been located in other genomes:\n
- \n",name);
}
printf("
- "); safef(extra,sizeof(extra),"%s=full",tdb->track); linkToOtherBrowserExtra(db, chrom, beg, end, extra); printf("%s \n",strSwapChar(title,'_',' ')); sqlFreeResult(&sr); } hFreeConn(&conn); freeMem(setting); if (gotOne) printf("
Currently the item \"%s\" has not been located in another genome.\n",name);
}
void mafPrettyOut(FILE *f, struct mafAli *maf, int lineSize,
boolean onlyDiff, int blockNo, struct hash *hash);
void doAtom( struct trackDb *tdb, char *item)
{
char table[HDB_MAX_TABLE_STRING];
boolean hasBin;
//struct bed *bed;
char query[512];
struct sqlResult *sr;
char **row;
//boolean firstTime = TRUE;
int start = cartInt(cart, "o");
//struct sqlConnection *conn = hAllocConn(database);
char *user = cfgOption("db.user");
char *password = cfgOption("db.password");
struct sqlConnection *sc;
struct atom ret;
genericHeader(tdb, item);
if (!hFindSplitTable(database, seqName, tdb->table, table, sizeof table, &hasBin))
errAbort("mafPrettyOut track %s not found", tdb->table);
#if 0
sqlSafef(query, sizeof query, "select * from %s where name = '%s' and chrom = '%s' and chromStart = %d", table, escapedName, seqName, start);
sr = sqlGetResult(conn, query);
printf("This is the item you clicked on:
\n");
while ((row = sqlNextRow(sr)) != NULL)
{
if (firstTime)
firstTime = FALSE;
else
htmlHorizontalLine();
bed = bedLoadN(row+hasBin, 4);
bedPrintPos(bed, 4, tdb);
}
sqlFreeResult(&sr);
sqlSafef(query, sizeof query, "select * from %s where name = '%s'", table, escapedName);
sr = sqlGetResult(conn, query);
while ((row = sqlNextRow(sr)) != NULL)
{
bed = bedLoadN(row+hasBin, 4);
if (bed->chromStart != start)
{
htmlHorizontalLine();
firstTime = FALSE;
printf("Another instances on %s:
\n",database);
bedPrintPos(bed, 4, tdb);
}
}
sqlFreeResult(&sr);
#endif
sc = sqlConnectRemote("localhost", user, password, "hgFixed");
sqlSafef(query, sizeof(query),
"select * from %s where name = '%s'", table, item);
sr = sqlGetResult(sc, query);
printf("Atom %s instances ('*' marks item you clicked on)
\n",item);
printf("
\n");
//printf("Ins#\tSpecies\t\tChrom\tStart\tEnd\tStrand\n");
printf( " # %-10s %-5s %12s %12s %10s %s %-10s %-10s\n",
"species","chrom", "start", "end", "length", "strand","fivePrime","threePrime");
while ((row = sqlNextRow(sr)) != NULL)
{
atomStaticLoad(row, &ret);
//atomOutput(&ret, stdout, '\t', '\n');
linkToOtherBrowser(ret.species, ret.chrom, ret.start, ret.end);
if (sameString(ret.chrom, seqName) && (start == ret.start) &&
sameString(ret.species, database))
printf("* ");
else
printf(" ");
printf( "%4d %-10s %-5s %12d %12d %10d %c %-10s %-10s\n",
ret.instance, ret.species,ret.chrom, ret.start + 1, ret.end,
ret.end - ret.start + 1, ret.strand[0],ret.fivePrime,ret.threePrime);
}
printf("");
sqlFreeResult(&sr);
if (!sameString("atom20080226d", table))
return;
printf("| ");
printf("Suh Trees \n"); printf("  ",item); printf(" |  ",item); printf(" |
|---|---|
| ");
printf("NJ Trees \n"); printf("  ",item); printf(" |  ",item); printf(" |
| ");
/*
printf("Gap UPGMA Trees \n"); printf("  ",item); printf(" |  ",item); printf(" |
Multiple Alignment Block %d of %d
", count, numBlocks); mafPrettyOut(stdout, mafAli, 70, FALSE, count++, NULL); if (mafAli->next != NULL) { struct mafAli *next = mafAli->next; struct mafComp *comp1 = mafAli->components; struct mafComp *comp2 = next->components; printf("
Gaps:\n"); for(; comp1 ; comp1 = comp1->next, comp2 = comp2->next) { int diff; char dbOnly[4096]; diff = comp2->start - (comp1->start + comp1->size); safef(dbOnly, sizeof(dbOnly), "%s", comp1->src); chopPrefix(dbOnly); printf("%-20s %d\n",hOrganism(dbOnly), diff); } printf("
"); } } } } char **getIdNameMap(struct trackDb *tdb, struct asColumn *col, int *size) /* Allocate and fill an array mapping id to name. Currently limited to specific columns. */ { char *idNameTable = trackDbSetting(tdb, "sourceTable"); if (!idNameTable || differentString("sourceIds", col->name)) return NULL; struct sqlResult *sr; char query[256]; char **row; char **idNames; sqlSafef(query, sizeof(query), "select max(id) from %s", idNameTable); struct sqlConnection *conn = hAllocConnTrack(database, tdb); int maxId = sqlQuickNum(conn, query); AllocArray(idNames, maxId+1); sqlSafef(query, sizeof(query), "select id, name from %s", idNameTable); sr = sqlGetResult(conn, query); int id; while ((row = sqlNextRow(sr)) != NULL) { id = sqlUnsigned(row[0]); if (id > maxId) errAbort("Internal error: id %d > maxId %d in %s", id, maxId, idNameTable); idNames[id] = cloneString(row[1]); } sqlFreeResult(&sr); hFreeConn(&conn); if (size) *size = maxId+1; return idNames; } void printIdOrLinks(struct asColumn *col, struct hash *fieldToUrl, struct trackDb *tdb, char *idList) /* if trackDb does not contain a "urls" entry for current column name, just print idList as it is. * Otherwise treat idList as a comma-sep list of IDs and print one row per id, with a link to url, * ($$ in url is OK, wildcards like $P, $p, are also OK) * */ { // try to find a fieldName=url setting in the "urls" tdb statement, print id if not found char *url = NULL; if (fieldToUrl != NULL) url = (char*)hashFindVal(fieldToUrl, col->name); if (url == NULL) { printf("
", fieldName, fieldVal, count); } slReverse(extraFields); return extraFields; } struct slPair *getFields(struct trackDb *tdb, char **row) /* return field names and their values as a list of slPairs. */ // TODO: refactor with getExtraFields { struct asObject *as = asForDb(tdb, database); if (as == NULL) return NULL; int fieldCount = slCount(as->columnList); struct slPair *fields = NULL; struct asColumn *col = as->columnList; int count = 0; for (count = 0; col != NULL && count < fieldCount; col = col->next) { struct slPair *field; AllocVar(field); char *fieldName = col->name; char *fieldVal = row[count]; field->name = fieldName; field->val = fieldVal; slAddHead(&fields, field); count++; } slReverse(fields); return fields; } void printEmbeddedTable(struct trackDb *tdb, struct embeddedTbl *thisTbl, struct dyString *dy) // Pretty print a '|' and ';' encoded table or a JSON encoded table from a bigBed field { jsIncludeFile("hgc.js", NULL); if (isNotEmpty(thisTbl->encodedTbl)) { if (startsWith("_json", thisTbl->field) || startsWith("json", thisTbl->field)) { struct jsonElement *jsElem = NULL; struct errCatch *errCatch = errCatchNew(); if (errCatchStart(errCatch)) jsElem = jsonParse(thisTbl->encodedTbl); errCatchEnd(errCatch); if (errCatch->gotError) warn("ERROR: JSON field '%s' for track '%s' is malformed: %s", thisTbl->field, tdb->track, errCatch->message->string); else if (errCatch->gotWarning) warn("Warning: %s", errCatch->message->string); errCatchFree(&errCatch); if (jsElem != NULL) { char *labelStr = thisTbl->title != NULL ? jsonStringEscape(thisTbl->title) : jsonStringEscape(thisTbl->field); dyStringPrintf(dy, "{label: \"%s\", data: %s},", labelStr, thisTbl->encodedTbl); } } else { printf("
| "); char table[4096]; safef(table, sizeof(table), "%s", thisTbl->encodedTbl); int swapped = strSwapStrs(table, 4096, ";", " | |
| "); if (swapped == -1) errAbort("Error substituting ';' for ' | |
| "); if (swapped == -1) errAbort("Error substituting '|' for ' | ' in hgc.c:printEmbeddedTable()"); printf("%s |
%s
\n", description); printf("| "); char *toPrint = replaceChars(line, "\t", " | "); printf("%s", toPrint); printf(" |
%s", afterPipe); } puts("
\n
| \n"); else if (col->isList || col->isArray || col->lowType->stringy || asTypesIsInt(col->lowType->type)) printIdOrLinks(col, fieldToUrl, tdb, fields[ix]); else if (asTypesIsFloating(col->lowType->type)) { double valDouble = strtod(fields[ix],NULL); if (errno == 0 && valDouble != 0) printf(" | %g | \n", valDouble); else printf("%s | \n", hubEncode(tdb, fields[ix])); // decided not to print error } else printf("%s | \n", hubEncode(tdb, fields[ix])); printCount++; } if (skipIds) slFreeList(skipIds); if (sepFields) slFreeList(sepFields); if (embeddedTblFields) { printf("
Sequence 1: %s
Sequence 2: %s
\n",row[hasBin+6], row[hasBin+7]); printCompareGenomeLinks(tdb,bed->name); } sqlFreeResult(&sr); getBedTopScorers(conn, tdb, table, item, start, bedSize); } printItemDetailsHtml(tdb, item); } #define INTRON 10 #define CODINGA 11 #define CODINGB 12 #define UTR5 13 #define UTR3 14 #define STARTCODON 15 #define STOPCODON 16 #define SPLICESITE 17 #define NONCONSPLICE 18 #define INFRAMESTOP 19 #define INTERGENIC 20 #define REGULATORY 21 #define LABEL 22 #define RED 0xFF0000 #define GREEN 0x00FF00 #define LTGREEN 0x33FF33 #define BLUE 0x0000FF #define MEDBLUE 0x6699FF #define PURPLE 0x9900cc #define BLACK 0x000000 #define CYAN 0x00FFFF #define ORANGE 0xDD6600 #define BROWN 0x663300 #define YELLOW 0xFFFF00 #define MAGENTA 0xFF00FF #define GRAY 0xcccccc #define LTGRAY 0x999999 #define WHITE 0xFFFFFF int setAttributeColor(int class) { switch (class) { case STARTCODON: return GREEN; case STOPCODON: return RED; case CODINGA: return MEDBLUE; case CODINGB: return PURPLE; case UTR5: case UTR3: return ORANGE; case INTRON: return LTGRAY; case SPLICESITE: case NONCONSPLICE: return BLACK; case INFRAMESTOP: return MAGENTA; case REGULATORY: return YELLOW; case INTERGENIC: return GRAY; case LABEL: default: return BLACK; } } void startColorStr(struct dyString *dy, int color, int track) { currentColor[track] = color; if (prevColor[track] != currentColor[track]) dyStringPrintf(dy,"",color); } void stopColorStr(struct dyString *dy, int track) { prevColor[track] = currentColor[track]; } void addTag(struct dyString *dy, struct dyString *tag) { dyStringPrintf(dy,"",tag->string); } void setClassStr(struct dyString *dy, int class, int track) { if (class == STARTCODON) dyStringAppend(dy,""); startColorStr(dy,setAttributeColor(class),track); } void resetClassStr(struct dyString *dy, int track) { stopColorStr(dy,track); } boolean isBlue(char *s) /* check for to see if this is colored blue (coding region)*/ { /* check for blue */ if (strstr(s,"6699FF") == NULL) return FALSE; else return TRUE; } int numberOfGaps(char *q,int size) /* count number of gaps in a string array */ { int i; int count = 0; for (i = 0 ; i
",qPtr, nextStart+1,nextEnd,nextEndIndex+1, tPtr,prevEnd, tPtr-nextStart-70, tPtr-(prevEnd+70)); else printf(" %d nextExon=%d-%d xon %d t %d prevEnd %d diffs %d %d
",qPtr, nextStart+1,nextEnd,nextEndIndex, tPtr, prevEnd, tPtr-nextStart-70, tPtr-(prevEnd+70)); #endif /* write out alignment, unless we are deep inside an intron */ if (tClass != INTRON || (tClass == INTRON && tPtr < nextStart-LINESIZE && tPtr< (prevEnd + posStrand ? LINESIZE : -LINESIZE))) { intronTruncated = 0; fputs(dyTprot->string,f); fputs(dyT->string,f); for (i=0; i
"); intronTruncated = TRUE; } } /* look for end of line */ if (oneSize > lineSize) oneSize = lineSize; sizeLeft -= oneSize; q += oneSize; t += oneSize; dyStringFree(&dyT); dyStringFree(&dyQ); dyStringFree(&dyQprot); dyStringFree(&dyTprot); } } } struct axt *getAxtListForGene(struct genePred *gp, char *nib, char *fromDb, char *toDb, struct lineFile *lf) /* get all axts for a gene */ { struct axt *axt, *axtGap; struct axt *axtList = NULL; int prevEnd = gp->txStart; // int prevStart = gp->txEnd; unused variable int tmp; while ((axt = axtRead(lf)) != NULL) { if (sameString(gp->chrom, axt->tName) && ( ( (axt->tStart <= gp->cdsStart && axt->tEnd >= gp->cdsStart) || (axt->tStart <= gp->cdsEnd && axt->tEnd >= gp->cdsEnd ) ) || ( axt->tStart < gp->cdsEnd && axt->tEnd > gp->cdsStart ) ) ) { if (gp->strand[0] == '-') { reverseComplement(axt->qSym, axt->symCount); reverseComplement(axt->tSym, axt->symCount); tmp = hChromSize(fromDb, axt->qName) - axt->qStart; axt->qStart = hChromSize(fromDb, axt->qName) - axt->qEnd; axt->qEnd = tmp; if (prevEnd < (axt->tStart)-1) { axtGap = createAxtGap(nib,gp->chrom,prevEnd,(axt->tStart),gp->strand[0]); reverseComplement(axtGap->qSym, axtGap->symCount); reverseComplement(axtGap->tSym, axtGap->symCount); slAddHead(&axtList, axtGap); } } else if (prevEnd < (axt->tStart)) { axtGap = createAxtGap(nib,gp->chrom,prevEnd,(axt->tStart),gp->strand[0]); slAddHead(&axtList, axtGap); } slAddHead(&axtList, axt); prevEnd = axt->tEnd; // prevStart = axt->tStart; unused variable } if (sameString(gp->chrom, axt->tName) && (axt->tStart > gp->txEnd)) { if ((prevEnd < axt->tStart) && prevEnd < min(gp->txEnd, axt->tStart)) { axtGap = createAxtGap(nib,gp->chrom,prevEnd,min(axt->tStart,gp->txEnd),gp->strand[0]); if (gp->strand[0] == '-') { reverseComplement(axtGap->qSym, axtGap->symCount); reverseComplement(axtGap->tSym, axtGap->symCount); } slAddHead(&axtList, axtGap); } else if (axtList == NULL) { axtGap = createAxtGap(nib,gp->chrom,prevEnd,gp->txEnd,gp->strand[0]); if (gp->strand[0] == '-') { reverseComplement(axtGap->qSym, axtGap->symCount); reverseComplement(axtGap->tSym, axtGap->symCount); } slAddHead(&axtList, axtGap); } break; } } if (gp->strand[0] == '+') slReverse(&axtList); return axtList ; } struct axt *getAxtListForRange(struct genePred *gp, char *nib, char *fromDb, char *toDb, char *alignment, char *qChrom, int qStart, int qEnd) /* get all axts for a chain */ { struct lineFile *lf ; struct axt *axt, *axtGap; struct axt *axtList = NULL; int prevEnd = gp->txStart; // int prevStart = gp->txEnd; unused variable int tmp; lf = lineFileOpen(getAxtFileName(gp->chrom, toDb, alignment, fromDb), TRUE); printf("file %s\n",lf->fileName); while ((axt = axtRead(lf)) != NULL) { /* if (sameString(gp->chrom , axt->tName)) * printf("axt %s qstart %d axt tStart %d\n",axt->qName, axt->qStart,axt->tStart); */ if ( sameString(gp->chrom, axt->tName) && sameString( qChrom, axt->qName) && positiveRangeIntersection( qStart, qEnd, axt->qStart, axt->qEnd) && positiveRangeIntersection(gp->txStart, gp->txEnd, axt->tStart, axt->tEnd) ) { if (gp->strand[0] == '-') { reverseComplement(axt->qSym, axt->symCount); reverseComplement(axt->tSym, axt->symCount); tmp = hChromSize(fromDb, axt->qName) - axt->qStart; axt->qStart = hChromSize(fromDb, axt->qName) - axt->qEnd; axt->qEnd = tmp; if (prevEnd < (axt->tStart)-1) { axtGap = createAxtGap(nib,gp->chrom,prevEnd,(axt->tStart)-1,gp->strand[0]); reverseComplement(axtGap->qSym, axtGap->symCount); reverseComplement(axtGap->tSym, axtGap->symCount); slAddHead(&axtList, axtGap); } } else if (prevEnd < (axt->tStart)-1) { axtGap = createAxtGap(nib,gp->chrom,prevEnd,(axt->tStart)-1,gp->strand[0]); slAddHead(&axtList, axtGap); } slAddHead(&axtList, axt); prevEnd = axt->tEnd; // prevStart = axt->tStart; unused variable } if (sameString(gp->chrom, axt->tName) && (axt->tStart > gp->txEnd+20000)) { if (axt->tStart > prevEnd) { axtGap = createAxtGap(nib,gp->chrom,prevEnd+1,(axt->tStart)-1,gp->strand[0]); if (gp->strand[0] == '-') { reverseComplement(axtGap->qSym, axtGap->symCount); reverseComplement(axtGap->tSym, axtGap->symCount); } slAddHead(&axtList, axtGap); } break; } } if (gp->strand[0] == '+') slReverse(&axtList); return axtList ; } void printCdsStatus(enum cdsStatus cdsStatus) /* print a description of a genePred cds status */ { switch (cdsStatus) { case cdsNone: /* "none" - No CDS (non-coding) */ printf("none (non-coding)
\n"); break; case cdsUnknown: /* "unk" - CDS is unknown (coding, but not known) */ printf("unknown (coding, but not known)
\n"); break; case cdsIncomplete: /* "incmpl" - CDS is not complete at this end */ printf("not complete
\n"); break; case cdsComplete: /* "cmpl" - CDS is complete at this end */ printf("complete
\n"); break; } } void showGenePos(char *name, struct trackDb *tdb) /* Show gene prediction position and other info. */ { char query[512]; char *liftDb = cloneString(trackDbSetting(tdb, "quickLiftDb")); char *db = (liftDb == NULL) ? database : liftDb; // tdb->table carries the hub_NNN_ prefix for quickLifted tracks; the // actual SQL table lives in the source assembly under the bare name. char *rootTable = (liftDb == NULL) ? tdb->table : trackHubSkipHubName(tdb->table); struct sqlConnection *conn = hAllocConn(db); struct genePred *gpList = NULL, *gp = NULL; char table[HDB_MAX_TABLE_STRING]; struct sqlResult *sr = NULL; char **row = NULL; char *classTable = trackDbSetting(tdb, GENEPRED_CLASS_TBL); if (!hFindSplitTable(db, seqName, rootTable, table, sizeof table, NULL)) errAbort("showGenePos track %s not found", rootTable); sqlSafef(query, sizeof(query), "name = \"%s\"", name); if (liftDb != NULL) { // Fetch via quickLiftSql so we get both the items and the set of // swapped chains that map them back to the destination assembly, then // lift the genePreds with calcLiftOverGenePreds. seqName/winStart/ // winEnd are destination coords; quickLiftSql picks up the chains // that overlap that window on the destination side. char *quickLiftFile = trackDbSetting(tdb, "quickLiftUrl"); struct hash *chainHash = newHash(8); char extraWhere[512]; sqlSafef(extraWhere, sizeof extraWhere, "name = \"%s\"", name); gpList = quickLiftGenePreds(conn, quickLiftFile, rootTable, seqName, winStart, winEnd, extraWhere, chainHash); calcLiftOverGenePreds(gpList, chainHash, 0.0, 0.0, TRUE, NULL, NULL, TRUE, FALSE); } else gpList = genePredReaderLoadQuery(conn, table, query); for (gp = gpList; gp != NULL; gp = gp->next) { printPos(gp->chrom, gp->txStart, gp->txEnd, gp->strand, FALSE, NULL); if(sameString(tdb->type,"genePred") && startsWith("ENCODE Gencode",tdb->longLabel) && startsWith("ENST",name)) { char *ensemblIdUrl = trackDbSetting(tdb, "ensemblIdUrl"); printf("Ensembl Transcript Id: "); if (ensemblIdUrl != NULL) printf("%s
", ensemblIdUrl,name,name); else printf("%s
",name); } if (gp->name2 != NULL && strlen(trimSpaces(gp->name2))> 0) { /* in Ensembl gene info downloaded from ftp site, sometimes the name2 field is populated with "noXref" because there is no alternate name. Replace this with "none" */ printf("Gene Symbol:"); if ((strlen(gp->name2) < 1) || (sameString(gp->name2, "noXref"))) printf(" none
\n"); else printf(" %s
\n",gp->name2); } char *ensemblSource = NULL; if (sameString("ensGene", table)) { if (hTableExists(db, "ensemblSource")) { sqlSafef(query, sizeof(query), "select source from ensemblSource where name='%s'", name); ensemblSource = sqlQuickString(conn, query); } } if ((gp->exonFrames != NULL) && (!genbankIsRefSeqNonCodingMRnaAcc(gp->name))) { if (ensemblSource && differentString("protein_coding",ensemblSource)) { printf("CDS Start: none (non-coding)
\n"); printf("CDS End: none (non-coding)
\n"); } else { printf("CDS Start: "); printCdsStatus((gp->strand[0] == '+') ? gp->cdsStartStat : gp->cdsEndStat); printf("CDS End: "); printCdsStatus((gp->strand[0] == '+') ? gp->cdsEndStat : gp->cdsStartStat); } } /* if a gene class table exists, get gene class and print */ if ((liftDb == NULL) && (classTable != NULL)) { if (hTableExists(db, classTable)) { sqlSafef(query, sizeof(query), "select class from %s where name = \"%s\"", classTable, name); sr = sqlGetResult(conn, query); /* print class */ if ((row = sqlNextRow(sr)) != NULL) printf("Prediction Class: %s
\n", row[0]); sqlFreeResult(&sr); if (sqlFieldIndex(conn, classTable, "level") > 0 ) { sqlSafef(query, sizeof(query), "select level from %s where name = \"%s\"", classTable, name); sr = sqlGetResult(conn, query); if ((row = sqlNextRow(sr)) != NULL) printf("Level: %s
\n", row[0]); sqlFreeResult(&sr); } if (sqlFieldIndex(conn, classTable, "transcriptType") > 0 ) { sqlSafef(query, sizeof(query), "select transcriptType from %s where name = \"%s\"", classTable, name); sr = sqlGetResult(conn, query); if ((row = sqlNextRow(sr)) != NULL) printf("Transcript type: %s
\n", row[0]); sqlFreeResult(&sr); } if (sqlFieldIndex(conn, classTable, "geneDesc") > 0 ) { sqlSafef(query, sizeof(query), "select geneDesc from %s where name = \"%s\"", classTable, name); sr = sqlGetResult(conn, query); if ((row = sqlNextRow(sr)) != NULL) if (differentString("NULL",row[0])) printf("Gene Description : %s
\n", row[0]); sqlFreeResult(&sr); } if (sqlFieldIndex(conn, classTable, "type") > 0 ) { sqlSafef(query, sizeof(query), "select type from %s where name = \"%s\"", classTable, name); sr = sqlGetResult(conn, query); if ((row = sqlNextRow(sr)) != NULL) if (differentString("NULL",row[0])) printf("Gene Type : %s
\n", row[0]); } } } if (gp->next != NULL) printf("
"); } genePredFreeList(&gpList); sqlFreeResult(&sr); hFreeConn(&conn); } void showGenePosMouse(char *name, struct trackDb *tdb, struct sqlConnection *connMm) /* Show gene prediction position and other info. */ { char query[512]; char *rootTable = tdb->table; struct sqlResult *sr; char **row; struct genePred *gp = NULL; boolean hasBin; int posCount = 0; char table[HDB_MAX_TABLE_STRING]; if (!hFindSplitTable(database, seqName, rootTable, table, sizeof table, &hasBin)) errAbort("showGenePosMouse track %s not found", rootTable); sqlSafef(query, sizeof query, "select * from %s where name = '%s'", table, name); sr = sqlGetResult(connMm, query); while ((row = sqlNextRow(sr)) != NULL) { if (posCount > 0) printf("
\n"); ++posCount; gp = genePredLoad(row + hasBin); printPos(gp->chrom, gp->txStart, gp->txEnd, gp->strand, FALSE, NULL); genePredFree(&gp); } sqlFreeResult(&sr); } void linkToPal(char *track, char *chrom, int start, int end, char *geneName) /* Make anchor tag to open pal window */ { printf("", geneName, hgPalName(), track, geneName, chrom, start, end); printf("CDS FASTA alignment from multiple alignment"); } void addPalLink(struct sqlConnection *conn, char *track, char *chrom, int start, int end, char *geneName) { struct slName *list = hTrackTablesOfType(conn, "wigMaf%%"); if (list != NULL) { puts("
\n", buffer, ya->yaleId); } sqlFreeResult(&sr); hFreeConn(&conn); } } printf("
Links to sequence:
\n"); printf("- \n");
if ((pepTable != NULL) && (pepName != NULL) && hGenBankHaveSeq(srcDb, pepName, pepTable))
{
puts("
- \n"); hgcAnchorSomewhere(pepClick, pepName, pepTable, seqName); printf("Predicted Protein
Links to sequence:
\n"); printf("- \n");
puts("
- \n"); hgcAnchorSomewhere(genomicClick, geneName, geneTable, seqName); printf("Genomic Sequence from assembly\n"); puts(" \n"); printf("
Links to sequence:
\n"); printf("- \n");
if (pepTable != NULL && hTableExists(database, pepTable))
{
hgcAnchorSomewhereDb(pepClick, pepName, pepTable, seqName, mousedb);
printf("
- Translated Protein \n"); } hgcAnchorSomewhereDb(mrnaClick, geneName, geneTrack, seqName, mousedb); printf("
- %s\n", mrnaDescription); hgcAnchorSomewhereDb(genomicClick, geneName, geneTrack, seqName, mousedb); printf("
- Genomic Sequence DNA sequence from assembly\n"); printf("
\n", x->oldId); printf("New ID: %s
\n", naForEmpty(x->newId)); printf("Old/New Mapping: %s
\n", x->status); if (x->note[0] != 0) printf("Notes: %s
\n", x->note); printf("
\n"); } sqlFreeResult(&sr); } geneShowCommon(item, tdb, pepTable); printItemDetailsHtml(tdb, item); } void pslDumpHtml(struct psl *pslList) /* print out psl header and data */ { struct psl* psl; printf("
\n");
printf("#match\tmisMatches\trepMatches\tnCount\tqNumInsert\tqBaseInsert\ttNumInsert\tBaseInsert\tstrand\tqName\tqSize\tqStart\tqEnd\ttName\ttSize\ttStart\ttEnd\tblockCount\tblockSizes\tqStarts\ttStarts\n");
for (psl = pslList; psl != NULL; psl = psl->next)
{
pslTabOut(psl, stdout);
}
printf("\n");
}
struct twoBitFile *getTwoBitFileFromUrlInSettings(struct trackDb *tdb, char *settingName)
/* Assume settingName is a URL to a two bit file and try and return it */
{
struct twoBitFile *tbf = NULL;
char *url = trackDbSetting(tdb, settingName);
if (url != NULL)
tbf = twoBitOpen(url);
return tbf;
}
struct twoBitFile *getOtherTwoBitUrl(struct trackDb *tdb)
/* Return open two bit file if otherTwoBitUrl setting is present */
{
return getTwoBitFileFromUrlInSettings(tdb, "otherTwoBitUrl");
}
struct psl *getPslAndSeq(struct trackDb *tdb, char *chromName, struct bigBedInterval *bb,
unsigned seqTypeField, DNA **retSeq, char **retCdsStr)
/* Read in psl and query sequence out of a bbiInverval on a give chromosome */
{
struct psl *psl= pslFromBigPsl(chromName, bb, seqTypeField, retSeq, retCdsStr);
DNA *dna = *retSeq;
if (dna == NULL)
{
struct twoBitFile *otherTbf = getOtherTwoBitUrl(tdb);
struct dnaSeq *seq = NULL;
if (otherTbf)
{
seq = twoBitReadSeqFrag(otherTbf, psl->qName, 0, 0);
if (seq)
*retSeq = dna = seq->dna;
}
}
return psl;
}
void genericBigPslClick(struct sqlConnection *conn, struct trackDb *tdb,
char *item, int start, int end)
/* Handle click in big psl track. */
{
struct psl* pslList = NULL;
char *fileName = bbiNameFromSettingOrTable(tdb, conn, tdb->table);
struct bbiFile *bbi = bigBedFileOpenAlias(fileName, chromAliasFindAliases);
struct lm *lm = lmInit(0);
int ivStart = start, ivEnd = end;
if (start == end)
{
// item is an insertion; expand the search range from 0 bases to 2 so we catch it:
ivStart = max(0, start-1);
ivEnd++;
}
if (cfgOptionBooleanDefault("drawDot", FALSE))
bigPslDotPlot(tdb, bbi, seqName, winStart, winEnd);
boolean showEvery = sameString(item, "PrintAllSequences");
boolean showAll = trackDbSettingOn(tdb, "showAll");
unsigned seqTypeField = bbExtraFieldIndex(bbi, "seqType");
struct bigBedInterval *bb, *bbList = NULL;
struct hash *chainHash = NULL;
struct hash *mapPsls = NULL; // mapping alignments quickLift reuses across items
char *quickLiftFile = trackDbSetting(tdb, "quickLiftUrl");
// A quickLifted track can only show what the chains around this window reach, so it takes
// the windowed query even when the track asks for every alignment of the item. The file
// holds the other assembly's alignments, so the window has to be turned into that
// assembly's coordinates before the query. quickLiftGetIntervals also hands back the
// chains needed to bring the alignments the other way.
// If showAll is on, show all alignments with this qName, not just the
// selected one.
if (quickLiftFile != NULL)
bbList = quickLiftGetIntervals(quickLiftFile, bbi, seqName, ivStart, ivEnd, &chainHash);
else if (showEvery)
{
struct bbiChromInfo *chrom, *chromList = bbiChromList(bbi);
for (chrom = chromList; chrom != NULL; chrom = chrom->next)
{
char *chromName = chrom->name;
int start = 0, end = chrom->size;
int itemsLeft = 0; // Zero actually means no limit....
struct bigBedInterval *intervalList = bigBedIntervalQuery(bbi, chromName,
start, end, itemsLeft, lm);
slCat(&bbList, intervalList);
}
}
else if (showAll)
{
int fieldIx;
struct bptFile *bpt = bigBedOpenExtraIndex(bbi, "name", &fieldIx);
bbList = bigBedNameQuery(bbi, bpt, fieldIx, item, lm);
}
else
bbList = bigBedIntervalQuery(bbi, seqName, ivStart, ivEnd, 0, lm);
/* print out extra fields */
for (bb = bbList; bb != NULL; bb = bb->next)
{
char *restFields[256];
int restCount = chopTabs(cloneString(bb->rest), restFields);
if (sameString(restFields[0], item))
{
/* print standard position information */
char *strand = restFields[2];
int bedSize = 25;
int restBedFields = bedSize - 3;
struct slPair *extraFieldPairs = NULL;
if (restCount > restBedFields)
{
char **extraFields = (restFields + restBedFields);
int extraFieldCount = restCount - restBedFields;
extraFieldPairs = getExtraFields(tdb, extraFields, extraFieldCount);
char *itemForUrl = getIdInUrl(tdb, item);
printCustomUrlWithFields(tdb, item, item, item == itemForUrl, extraFieldPairs);
if (itemForUrl)
printIframe(tdb, itemForUrl);
printPos(seqName, ivStart, ivEnd, strand, FALSE, item);
int printCount = extraFieldsPrint(tdb,NULL,extraFields, extraFieldCount);
printCount += 0;
}
}
}
char *bedRow[32];
char startBuf[16], endBuf[16];
int lastChromId = -1;
char chromName[bbi->chromBpt->keySize+1];
boolean firstTime = TRUE;
struct hash *seqHash = hashNew(0);
struct dyString *sequencesText = dyStringNew(256);
int sequencesFound = 0;
for (bb = bbList; bb != NULL; bb = bb->next)
{
bbiCachedChromLookup(bbi, bb->chromId, lastChromId, chromName, sizeof(chromName));
lastChromId=bb->chromId;
bigBedIntervalToRow(bb, chromName, startBuf, endBuf, bedRow, 4);
if (showEvery || sameString(bedRow[3], item))
{
char *cdsStr, *seq;
struct psl *psl= getPslAndSeq(tdb, chromName, bb, seqTypeField, &seq, &cdsStr);
if (chainHash != NULL)
{
struct psl *lifted = quickLiftPsl(chainHash, &mapPsls, psl);
pslFree(&psl);
if (lifted == NULL)
continue; // nothing in the chains places this alignment
psl = lifted;
}
slAddHead(&pslList, psl);
// we're assuming that if there are multiple psl's with the same id that
// they are the same query sequence so we only put out one set of sequences
if (!hashLookup(seqHash, bedRow[3]) && !isEmpty(seq)) // if there is a query sequence
{
if (firstTime)
{
firstTime = FALSE;
printf("Links to sequence:
\n"); printf("- \n");
}
if (!isEmpty(cdsStr)) // if we have CDS
{
puts("
- \n"); hgcAnchorSomewhereExt("htcTranslatedBigPsl", bedRow[3], "translate", bedRow[0], sqlUnsigned(bedRow[1]), sqlUnsigned(bedRow[2]), tdb->track); if (showEvery) printf("Translated Amino Acid Sequence from Query Sequence %s\n", bedRow[3]); else printf("Translated Amino Acid Sequence from Query Sequence\n"); puts(" \n"); } if (psl->qSize <= MAX_DISPLAY_QUERY_SEQ_SIZE) { puts("
- \n"); hgcAnchorSomewhereExt("htcBigPslSequence", bedRow[3], tdb->track, bedRow[0], sqlUnsigned(bedRow[1]), sqlUnsigned(bedRow[2]), tdb->track); if (showEvery) printf("Query Sequence %s \n", bedRow[3]); else printf("Query Sequence \n"); puts(" \n"); } hashAdd(seqHash, bedRow[3], ""); dyStringPrintf(sequencesText, ">%s\n%s\n\n", bedRow[3], seq); ++sequencesFound; } } if (!showEvery && !firstTime) break; } if (!firstTime) printf("
Genomic Alignments
"); else printf("%s/Genomic Alignments
", item); /* Hub track names have special characters replaced with underbar, but tdb->table does not. */ char *aliTable = cloneString(tdb->table); if (isHubTrack(aliTable)) trackHubFixName(aliTable); // pslIsProtein reads through its argument, and the list is empty whenever nothing in the // window matched the item, or, on a quickLifted track, nothing in the window could be // lifted. if (showEvery || ((pslList != NULL) && pslIsProtein(pslList))) printAlignmentsSimple(pslList, start, "htcBigPslAli", aliTable, item); else printAlignmentsExtra(pslList, start, "htcBigPslAli", "htcBigPslAliInWindow", aliTable, item); freeMem(aliTable); pslFreeList(&pslList); if (showEvery && sequencesFound > 0) { printf("\n"); printf("Input Sequences:
\n"); printf("\n"); dyStringFree(&sequencesText); } printItemDetailsHtml(tdb, item); } void genericPslClick(struct sqlConnection *conn, struct trackDb *tdb, char *item, int start, char *subType) /* Handle click in generic psl track. */ { struct psl* pslList = getAlignments(conn, tdb->table, item); // For a quickLifted track the alignments came out of the other assembly, and so did the // sequence the check below looks for, so both have to name that assembly. Move the // alignments onto the reference before anything prints a position. Only the ones the // chains around this window can place survive, which leaves out alignments of the same // accession elsewhere in the genome. char *liftDb = trackDbSetting(tdb, "quickLiftDb"); char *srcDb = (liftDb != NULL) ? liftDb : database; char *quickLiftFile = trackDbSetting(tdb, "quickLiftUrl"); if ((quickLiftFile != NULL) && (pslList != NULL)) pslList = quickLiftPsls(quickLiftChainHash(quickLiftFile, seqName, winStart, winEnd), pslList); /* check if there is an alignment available for this sequence. This checks * both genbank sequences and other sequences in the seq table. If so, * set it up so they can click through to the alignment. */ if (hGenBankHaveSeq(srcDb, item, NULL)) { printf("
%s/Genomic Alignments
", item); if (sameString("protein", subType)) printAlignments(pslList, start, "htcProteinAli", tdb->table, item); else printAlignments(pslList, start, "htcCdnaAli", tdb->table, item); } else { /* just dump the psls */ pslDumpHtml(pslList); } pslFreeList(&pslList); printItemDetailsHtml(tdb, item); } static char *getParentTableName(struct trackDb *tdb) /* Get the track table or composite track parent table if applicable. */ { tdb = trackDbTopLevelSelfOrParent(tdb); return tdb->table; } static char *getParentTrackName(struct trackDb *tdb) /* Get the track name or composite track parent name if applicable. */ { tdb = trackDbTopLevelSelfOrParent(tdb); return tdb->track; } void printTBSchemaLink(struct trackDb *tdb) /* Make link to TB schema -- unless this is an on-the-fly (tableless) track. */ { if (hTableOrSplitExists(database, tdb->table)) { char *trackTable = getParentTableName(tdb); printf("\n", database, tdb->grp, trackTable, tdb->table, seqName, winStart+1, winEnd); } } void printTrackUiLink(struct trackDb *tdb) /* Make link to hgTrackUi. */ { char *trackName = getParentTrackName(tdb); struct trackDb *parentTdb = tdb; if (!sameString(trackName, tdb->track)) parentTdb = hTrackDbForTrack(database, trackName); // shortLabel comes from trackDb, which a track hub controls, escape it printf("\n", hTrackUiForTrack(tdb->track), database, trackName, cartSidUrlString(cart), htmlEncode(parentTdb->shortLabel)); } void printDataRestrictionDate(struct trackDb *tdb) /* If this annotation has a dateUnrestricted trackDb setting, print it */ { char *restrictionDate = encodeRestrictionDateDisplay(database,tdb); if (restrictionDate != NULL) { printf("Restricted until: %s\n", restrictionDate); freeMem(restrictionDate); } } static void printOrigAssembly(struct trackDb *tdb) /* If this annotation has been lifted, print the original * freeze, as indicated by the "origAssembly" trackDb setting */ { trackDbPrintOrigAssembly(tdb, database); } static char *getHtmlFromSelfOrParent(struct trackDb *tdb, char *liftDb) /* Get html from self or from parent if not in self. */ { for (;tdb != NULL; tdb = tdb->parent) { if (sameString(trackHubSkipHubName(tdb->track), "quickLiftChain")) tdb->html = hFileContentsOrWarning(hHelpFile(trackHubSkipHubName(tdb->track))); if (liftDb && isEmpty(tdb->html)) tdb->html = getTrackHtml(liftDb, tdb->table); if (tdb->html != NULL && tdb->html[0] != 0) { - // resolve $hgsid, which hgTrackDb had no cart to resolve, and for a hub the rest of - // its description page variables: a hub page never went through hgTrackDb at all. - // tdb is the track the html belongs to, so $parentTrack means its container. - hVarSubstTrackDbHtml(cart, tdb, database); + // a hub's description page never went through hgTrackDb, so resolve its variables + // here. tdb is the track the html belongs to, so $parentTrack means its container. + hVarSubstTrackDbHtml(tdb, database); return tdb->html; } } return NULL; } void printTrackHtml(struct trackDb *tdb) /* If there's some html associated with track print it out. Also print * last update time for data table and make a link * to the TB table schema page for this table. */ { // safety net: a few detail pages never print a position, so flush the iframe here printPendingIframe(); if (!isCustomTrack(tdb->track) && !isMyVariantsType(tdb->type)) { printRelatedTracks(database, trackHash, tdb, cart); extraUiLinks(database, tdb, cart); printTrackUiLink(tdb); printOrigAssembly(tdb); printDataVersion(database, tdb); printUpdateTime(database, tdb, NULL); printDataRestrictionDate(tdb); } char *liftDb = cloneString(trackDbSetting(tdb, "quickLiftDb")); char *html = getHtmlFromSelfOrParent(tdb, liftDb); if (html != NULL && html[0] != 0) { htmlHorizontalLine(); // Add pennantIcon printPennantIconNote(tdb); // Wrap description html in div with limited width, so when the page is very wide // due to long details, the user doesn't have to scroll right to read the description. puts("
");
puts(html);
puts("
");
}
hPrintf("\n"); +jsAddHgsidToLinks(); } static struct chain *quickLiftChainInRange(struct trackDb *tdb, int id) /* Load one chain out of the assembly the track came from and map it onto the reference. * Every chain in the window is loaded and then matched on id, rather than asking for the * one id: the chain's sequence name in the other assembly is not known here, and the * by-id loaders abort when the id is not in the range they were given. */ { char *liftDb = trackDbSetting(tdb, "quickLiftDb"); char *table = NULL; quickLiftResolveTable(tdb, trackHubSkipHubName(tdb->table), &table, &liftDb); char *quickLiftFile = trackDbSetting(tdb, "quickLiftUrl"); char *chainFile = NULL, *linkFile = NULL; if (startsWith("big", tdb->type)) { chainFile = trackDbSetting(tdb, "bigDataUrl"); linkFile = trackDbSetting(tdb, "linkDataUrl"); if (linkFile == NULL) linkFile = bigChainGetLinkFile(chainFile); } struct hash *chainHash = newHash(8); struct hash *mapPsls = NULL; struct quickLiftRange *range, *rangeList = quickLiftSourceRanges(quickLiftFile, seqName, winStart, winEnd, chainHash); for (range = rangeList; range != NULL; range = range->next) { struct chain *chain, *chainList; if (chainFile != NULL) chainList = chainLoadIdRangeHub(NULL, chainFile, linkFile, range->chrom, range->start, range->end, -1); else chainList = chainLoadRange(liftDb, table, range->chrom, range->start, range->end); for (chain = chainList; chain != NULL; chain = chain->next) { if (chain->id != id) continue; struct chain *lifted = quickLiftChain(chainHash, &mapPsls, chain); if (lifted != NULL) return lifted; } } return NULL; } struct chain *chainLoadItemInRange(struct trackDb *tdb, char *item) /* Load up parts of chain that intersect seqName:winStart-winEnd */ { struct chain *chain = NULL; int id = sqlUnsigned(item); if (quickLiftIsLifted(tdb) && !quickLiftIsOwnChainTrack(tdb)) { chain = quickLiftChainInRange(tdb, id); if (chain == NULL) errAbort("Couldn't lift chain %d into %s:%d-%d", id, seqName, winStart, winEnd); return chain; } if (startsWith("big", tdb->type)) { char *fileName = trackDbSetting(tdb, "bigDataUrl"); char *linkFileName = trackDbSetting(tdb, "linkDataUrl"); if (linkFileName == NULL) { char *bigDataUrl = cloneString(trackDbSetting(tdb, "bigDataUrl")); char *dot = strrchr(bigDataUrl, '.'); if (dot == NULL) errAbort("No linkDataUrl in track %s", tdb->track); *dot = 0; char buffer[4096]; safef(buffer, sizeof buffer, "%s.link.bb", bigDataUrl); linkFileName = buffer; } chain = chainLoadIdRangeHub(database, fileName, linkFileName, seqName, winStart, winEnd, id); } else { chain = chainLoadIdRange(database, tdb->table, seqName, winStart, winEnd, id); } return chain; } void qChainRangePlusStrand(struct chain *chain, int *retQs, int *retQe) /* Return range of bases covered by chain on q side on the plus * strand. */ { if (chain == NULL) errAbort("Can't find range in null query chain."); if (chain->qStrand == '-') { *retQs = chain->qSize - chain->qEnd+1; *retQe = chain->qSize - chain->qStart; } else { *retQs = chain->qStart+1; *retQe = chain->qEnd; } } struct chain *chainDbLoad(struct sqlConnection *conn, char *db, char *track, char *chrom, int id) /* Load chain. */ { char table[HDB_MAX_TABLE_STRING]; char query[256]; struct sqlResult *sr; char **row; boolean hasBin; struct chain *chain; if (!hFindSplitTable(db, seqName, track, table, sizeof table, &hasBin)) errAbort("No %s track in database %s for %s", track, db, seqName); sqlSafef(query, sizeof(query), "select * from %s where id = %d", table, id); sr = sqlGetResult(conn, query); row = sqlNextRow(sr); if (row == NULL) errAbort("Can't find %d in %s", id, table); chain = chainHeadLoad(row + hasBin); sqlFreeResult(&sr); chainDbAddBlocks(chain, track, conn); return chain; } void linkToOtherBrowserHub(char *otherDb, char *chrom, int start, int end, char *hubUrl) /* Make anchor tag to open another browser window. */ { printf("", hubUrl); } void linkToOtherBrowserExtra(char *otherDb, char *chrom, int start, int end, char *extra) /* Make anchor tag to open another browser window. */ { printf("", otherDb, extra, chrom, start+1, end); } void linkToOtherBrowserSearch(char *otherDb, char *tag) /* Make anchor tag to open another browser window. */ { printf("", otherDb, tag); } void linkToOtherBrowser(char *otherDb, char *chrom, int start, int end) /* Make anchor tag to open another browser window. */ { printf("", otherDb, chrom, start+1, end); } void linkToOtherBrowserTitle(char *otherDb, char *chrom, int start, int end, char *title) /* Make anchor tag to open another browser window. */ { htmlPrintf("", otherDb, chrom, start+1, end); } void chainToOtherBrowser(struct chain *chain, char *otherDb, char *otherOrg, char *hubUrl) /* Put up link that lets us use chain to browser on * corresponding window of other species. */ { struct chain *subChain = NULL, *toFree = NULL; int qs,qe; chainSubsetOnT(chain, winStart, winEnd, &subChain, &toFree); if (subChain != NULL && otherOrg != NULL) { qChainRangePlusStrand(subChain, &qs, &qe); if (hubUrl) linkToOtherBrowserHub(otherDb, subChain->qName, qs-1, qe, hubUrl); else linkToOtherBrowser(otherDb, subChain->qName, qs-1, qe); htmlPrintf("Open %s browser at position corresponding to the part of chain that is in this window.
\n", trackHubSkipHubName(otherOrg)); } chainFree(&toFree); } struct dnaSeq *otherChromSeq(struct twoBitFile *otherTbf, char *otherDb, char *otherChrom, int otherStart, int otherEnd) /* This fetches sequence from tdb->otherTwoBitUrl if available, if not * tries to find it via otherDb */ { if (otherTbf != NULL) { return twoBitReadSeqFrag(otherTbf, otherChrom, otherStart, otherEnd); } else { return hChromSeq(otherDb, otherChrom, otherStart, otherEnd); } } void doSnakeChainClick(struct trackDb *tdb, char *itemName, char *otherDb) /* Put up page for chain snakes. */ { struct trackDb *parentTdb = trackDbTopLevelSelfOrParent(tdb); char *qName = cartOptionalString(cart, "qName"); int qs = atoi(cartOptionalString(cart, "qs")); int qe = atoi(cartOptionalString(cart, "qe")); int qWidth = atoi(cartOptionalString(cart, "qWidth")); char *aliasQName = qName; boolean otherIsActive = FALSE; char *hubUrl = NULL; if (hDbIsActive(otherDb)) otherIsActive = TRUE; else hubUrl = genarkUrl(otherDb); // may be NULL char headerText[256]; safef(headerText, sizeof headerText, "reference: %s, query: %s\n", trackHubSkipHubName(database), trackHubSkipHubName(otherDb) ); genericHeader(parentTdb, headerText); if (hubUrl != NULL) { printf("%s:%d-%d link to block in query assembly: %s
\n", otherDb, aliasQName, qs, qe, aliasQName, qs, qe, trackHubSkipHubName(otherDb)); } else if (otherIsActive) htmlPrintf("%s:%d-%d link to block in query assembly: %s
\n", otherDb, aliasQName, qs, qe, aliasQName, qs, qe, trackHubSkipHubName(otherDb)); int qCenter = (qs + qe) / 2; int newQs = qCenter - qWidth/2; int newQe = qCenter + qWidth/2; if (hubUrl != NULL) { printf("%s:%d-%d link to same window size in query assembly: %s
\n", otherDb, aliasQName, newQs, newQe,aliasQName, newQs, newQe, trackHubSkipHubName(otherDb) ); } else if (otherIsActive) htmlPrintf("%s:%d-%d link to same window size in query assembly: %s
\n", otherDb, aliasQName, newQs, newQe,aliasQName, newQs, newQe, trackHubSkipHubName(otherDb) ); printTrackHtml(tdb); } void genericChainClick(struct sqlConnection *conn, struct trackDb *tdb, char *item, int start, char *otherDb) /* Handle click in chain track, at least the basics. */ { boolean doSnake = cartOrTdbBoolean(cart, tdb, "doSnake" , FALSE) && cfgOptionBooleanDefault("canSnake", TRUE); if (doSnake) return doSnakeChainClick(tdb, item, otherDb); struct twoBitFile *otherTbf = getOtherTwoBitUrl(tdb); char *thisOrg = hOrganism(database); char *otherOrg = NULL; struct chain *chain = NULL, *subChain = NULL, *toFree = NULL; int chainWinSize; double subSetScore = 0.0; int qs, qe; boolean nullSubset = FALSE; boolean otherIsActive = FALSE; char *hubUrl = NULL; // if otherDb is a genark browser if (hDbIsActive(otherDb)) otherIsActive = TRUE; else hubUrl = genarkUrl(otherDb); // may be NULL if (! sameWord(otherDb, "seq")) { otherOrg = hOrganism(otherDb); } if (otherOrg == NULL) { /* use first word of chain label (count on org name as first word) */ otherOrg = firstWordInLine(cloneString(tdb->shortLabel)); } chain = chainLoadItemInRange(tdb, item); if (startsWith("big", tdb->type)) { if (!otherIsActive) // if this isn't a native database, check to see if it's a hub { struct trackHubGenome *genome = trackHubGetGenomeUndecorated(otherDb); if (genome) { otherIsActive = TRUE; otherDb = genome->name; } } } chainSubsetOnT(chain, winStart, winEnd, &subChain, &toFree); if (subChain == NULL) nullSubset = TRUE; else if (otherIsActive && subChain != chain) { char *linearGap = trackDbSettingOrDefault(tdb, "chainLinearGap", "loose"); struct gapCalc *gapCalc = gapCalcFromFile(linearGap); struct axtScoreScheme *scoreScheme = axtScoreSchemeDefault(); int qStart = subChain->qStart; int qEnd = subChain->qEnd ; struct dnaSeq *tSeq = hDnaFromSeq(database, subChain->tName, subChain->tStart, subChain->tEnd, dnaLower); struct dnaSeq *qSeq = NULL; char *matrix = trackDbSetting(tdb, "matrix"); if (matrix != NULL) { char *words[64]; int size = chopByWhite(matrix, words, 64) ; if (size == 2 && atoi(words[0]) == 16) { scoreScheme = axtScoreSchemeFromBlastzMatrix(words[1], 400, 30); } else { if (size != 2) errAbort("error parsing matrix entry in trackDb, expecting 2 word got %d ", size); else errAbort("error parsing matrix entry in trackDb, size 16 matrix, got %d ", atoi(words[0])); } } if (subChain->qStrand == '-') reverseIntRange(&qStart, &qEnd, subChain->qSize); qSeq = otherChromSeq(otherTbf, otherDb, subChain->qName, qStart, qEnd); if (subChain->qStrand == '-') reverseComplement(qSeq->dna, qSeq->size); subChain->score = chainCalcScoreSubChain(subChain, scoreScheme, gapCalc, qSeq, tSeq); subSetScore = subChain->score; } chainFree(&toFree); htmlPrintf("%s position: ", trackHubSkipHubName(thisOrg)); printf("%s:%d-%d size: %d
\n", database, chain->tName, chain->tStart+1, chain->tEnd, chain->tName, chain->tStart+1, chain->tEnd, chain->tEnd-chain->tStart); printf("Strand: %c
\n", chain->qStrand); qChainRangePlusStrand(chain, &qs, &qe); if (sameWord(otherDb, "seq")) { htmlPrintf("%s position: %s:%d-%d size: %d
\n", otherOrg, chain->qName, qs, qe, chain->qEnd - chain->qStart); } else { /* prints link to other db browser only if db exists and is active */ /* else just print position with no link for the other db */ htmlPrintf("%s position: ", otherOrg); if (otherIsActive) { printf(" ", otherDb, chain->qName, qs, qe); } else if (hubUrl != NULL) { printf(" ", chain->qName, qs, qe); } htmlPrintf("%s:%d-%d", chain->qName, qs, qe); if (otherIsActive || hubUrl) printf(""); printf(" size: %d
\n", chain->qEnd - chain->qStart); } htmlPrintf("Chain ID: %s
\n", item); printf("Score: %1.0f\n", chain->score); if (nullSubset) printf("Score within browser window: N/A (no aligned bases)
\n"); else if (otherIsActive && subChain != chain) printf(" Approximate Score within browser window: %1.0f
\n", subSetScore); else printf("
\n"); boolean normScoreAvailable = chainDbNormScoreAvailable(tdb); // The normalized score lives in the chain table, so for a quickLifted track it has to be // read from the assembly the chain came from. Against the assembly on screen the table // name either does not resolve, or resolves to a same-named table there and returns // somebody else's chain, which is worse. char *normDb = database; struct sqlConnection *normConn = conn; char *normTable = tdb->table; boolean lifted = quickLiftIsLifted(tdb) && !quickLiftIsOwnChainTrack(tdb); if (lifted) { normDb = trackDbSetting(tdb, "quickLiftDb"); normTable = trackHubSkipHubName(tdb->table); normConn = hAllocConn(normDb); } if (normScoreAvailable) { char tableName[HDB_MAX_TABLE_STRING]; if (!hFindSplitTable(normDb, chain->tName, normTable, tableName, sizeof tableName, NULL)) errAbort("genericChainClick track %s not found", normTable); char query[256]; struct sqlResult *sr; char **row; sqlSafef(query, ArraySize(query), "select normScore from %s where id = '%s'", tableName, item); sr = sqlGetResult(normConn, query); if ((row = sqlNextRow(sr)) != NULL) { double normScore = atof(row[0]); int basesAligned = chain->score / normScore; printf("Normalized Score: %1.0f (aligned bases: %d)", normScore, basesAligned); } sqlFreeResult(&sr); printf("
\n"); } if (lifted) hFreeConn(&normConn); if (quickLiftIsLifted(tdb) && !quickLiftIsOwnChainTrack(tdb)) // A lifted chain is only worked out over the window being viewed, so the whole chain's // extent is not knowable here and the usual sentence would be wrong. htmlPrintf("
This chain comes from %s and is mapped onto %s as the browser draws it, so " "the fields above describe the part of it around the window rather than the " "whole chain.
\n", trackDbSetting(tdb, "quickLiftDb"), trackHubSkipHubName(database)); else printf("
Fields above refer to entire chain or gap, not just the part inside the window.
\n"); printf("
\n"); chainWinSize = min(winEnd-winStart, chain->tEnd - chain->tStart); /* Show alignment if the database exists and */ /* if there is a chromInfo table for that database and the sequence */ /* file exists. This means that alignments can be shown on the archive */ /* server (or in other cases) if there is a database with a chromInfo table, */ /* the sequences are available and there is an entry added to dbDb for */ /* the otherDb. */ if (otherTbf != NULL || (!startsWith("big", tdb->type) && sqlDatabaseExists(otherDb) && chromSeqFileExists(otherDb, chain->qName))) { if (chainWinSize < 1000000) { printf("View "); hgcAnchorSomewhere("htcChainAli", item, tdb->track, chain->tName); printf("DNA sequence alignment details of parts of chain within browser " "window.
\n"); } else { printf("Zoom so that browser window covers 1,000,000 bases or less " "and return here to see alignment details.
\n"); } } if (!sameWord(otherDb, "seq") && (otherIsActive || hubUrl)) { chainToOtherBrowser(chain, otherDb, otherOrg, hubUrl); } /* if (!sameWord(otherDb, "seq") && (hDbIsActive(otherDb))) { chainToOtherBrowser(chain, otherDb, otherOrg); } */ chainFree(&chain); } char *trackTypeInfo(char *track) /* Return type info on track. You can freeMem result when done. */ { struct slName *trackDbs = hTrackDbList(), *oneTrackDb; struct sqlConnection *conn = hAllocConn(database); char buf[512]; char query[256]; for (oneTrackDb = trackDbs; oneTrackDb != NULL; oneTrackDb = oneTrackDb->next) { if (sqlTableExists(conn, oneTrackDb->name)) { sqlSafef(query, sizeof(query), "select type from %s where tableName = '%s'", oneTrackDb->name, track); if (sqlQuickQuery(conn, query, buf, sizeof(buf)) != NULL) break; } } if (oneTrackDb == NULL) errAbort("%s isn't in the trackDb from the hg.conf", track); slNameFreeList(&trackDbs); hFreeConn(&conn); return cloneString(buf); } void findNib(char *db, char *chrom, char nibFile[512]) /* Find nib file corresponding to chromosome in given database. */ { struct sqlConnection *conn = sqlConnect(db); char query[256]; sqlSafef(query, sizeof(query), "select fileName from chromInfo where chrom = '%s'", chrom); if (sqlQuickQuery(conn, query, nibFile, 512) == NULL) errAbort("Sequence %s isn't in database %s", chrom, db); sqlDisconnect(&conn); } struct dnaSeq *loadGenomePart(char *db, char *chrom, int start, int end) /* Load genomic dna from given database and position. */ { char nibFile[512]; findNib(db, chrom, nibFile); return hFetchSeq(nibFile, chrom, start, end); } void printLabeledNumber(char *org, char *label, long long number) /* Print label: in bold face, and number with commas. */ { char *space = " "; if (org == NULL) org = space = ""; htmlPrintf("%s%s%s: ", org, space, label); printLongWithCommas(stdout, number); printf("
\n"); } void printLabeledPercent(char *org, char *label, long p, long q) /* Print label: in bold, then p, and then 100 * p/q */ { char *space = " "; if (org == NULL) org = space = ""; htmlPrintf("%s%s%s: ", org, space, label); printLongWithCommas(stdout, p); if (q != 0) printf(" (%3.1f%%)", 100.0 * p / q); printf("
\n"); } static struct netAlign *bigNetLoadOne(struct trackDb *tdb, char *chrom, int start, unsigned level, boolean *retClipped) /* Load the record from a bigNet file at the given level that covers start. * Returns NULL if there isn't one. Sets *retClipped when a quickLifted row would * only lift with its ends pulled in, so the extent reported is the visible part * rather than the whole item. */ { char *fileName = hReplaceGbdb(trackDbSetting(tdb, "bigDataUrl")); char *quickLiftFile = trackDbSetting(tdb, "quickLiftUrl"); if (fileName == NULL) errAbort("No bigDataUrl in track %s", tdb->track); struct lm *lm = lmInit(0); struct bbiFile *bbi = bigBedFileOpenAlias(fileName, chromAliasFindAliases); struct hash *chainHash = NULL; struct bigBedInterval *bb, *bbList; struct netAlign *na = NULL; /* A quickLifted net is read over the whole window, because that is the lift the image * was drawn from, and each row is then tested against the clicked base once it has * landed. */ if (quickLiftFile != NULL) bbList = quickLiftGetIntervals(quickLiftFile, bbi, chrom, winStart, winEnd, &chainHash); else bbList = bigBedIntervalQuery(bbi, chrom, start, start+1, 0, lm); for (bb = bbList; bb != NULL; bb = bb->next) { struct bigNet bn; unsigned tStart = bb->start, tEnd = bb->end; boolean thisClipped = FALSE; if (quickLiftFile != NULL) { /* The unclipped lift first, so the page can report the item's whole extent. * An item too big for the chains loaded here lifts only with its ends pulled * in, and that is how the image drew it, so fall back to the clipped lift * rather than reporting an item the reader can plainly see as missing. */ struct bed *bed = quickLiftIntervalsToBed(bbi, chainHash, bb); if (bed == NULL) { bed = quickLiftIntervalsToBedClip(bbi, chainHash, bb); thisClipped = TRUE; } if (bed == NULL) continue; if (!sameString(bed->chrom, chrom)) { bedFree(&bed); continue; } tStart = bed->chromStart; tEnd = bed->chromEnd; bedFree(&bed); } if ((tStart > start) || (tEnd <= start)) continue; bigNetFromInterval(bbi, bb, fileName, &bn); if (bn.level != level) continue; if (retClipped != NULL) *retClipped = thisClipped; AllocVar(na); na->level = bn.level; na->tName = cloneString(chrom); na->tStart = tStart; na->tEnd = tEnd; safecpy(na->strand, sizeof na->strand, bn.strand); na->qName = cloneString(bn.name); na->qStart = bn.qStart; na->qEnd = bn.qEnd; na->chainId = bn.chainId; na->ali = bn.ali; na->score = bn.chainScore; na->qOver = bn.qOver; na->qFar = bn.qFar; na->qDup = bn.qDup; na->type = cloneString(bn.type); na->tN = bn.tN; na->qN = bn.qN; na->tR = bn.tR; na->qR = bn.qR; na->tNewR = bn.tNewR; na->qNewR = bn.qNewR; na->tOldR = bn.tOldR; na->qOldR = bn.qOldR; na->tTrf = bn.tTrf; na->qTrf = bn.qTrf; break; } bbiFileClose(&bbi); lmCleanup(&lm); return na; } static char *netChainTrackName(struct trackDb *tdb, char *chainTrack) /* The type line of a net track names its chain track without any hub prefix. * If the net track came from a hub, so did the chain track it names. */ { if (!isHubTrack(tdb->track)) return chainTrack; char buf[256]; safef(buf, sizeof buf, "hub_%d_%s", hubIdFromTrackName(tdb->track), chainTrack); return cloneString(buf); } static struct trackDb *netChainTdb(char *chainTrack) /* The tdb of the chain track a net track's type line names, or NULL if this assembly * has no such track. A quickLifted net is the case with none: the net comes across on * its own and its chain track stays behind on the source assembly. */ { return (trackHash == NULL) ? NULL : hashFindVal(trackHash, chainTrack); } void genericNetClick(struct sqlConnection *conn, struct trackDb *tdb, char *item, int start, char *otherDb, char *chainTrack) /* Generic click handler for net tracks. */ { char table[HDB_MAX_TABLE_STRING]; boolean hasBin; char query[256]; struct sqlResult *sr; char **row; struct netAlign *net; char *org = hOrganism(database); char *otherOrg = hOrganism(otherDb); char *otherOrgBrowser = otherOrg; int tSize, qSize; int netWinSize; struct chain *chain; boolean isBig = startsWith("big", tdb->type); /* A quickLifted net has moved to this assembly on its own. The chain track its type * line names is still on the source assembly, so there is no alignment to show and no * chain to follow -- only the net itself lifted. */ boolean isLifted = isBig && (trackDbSetting(tdb, "quickLiftUrl") != NULL); struct trackDb *chainTdb = NULL; /* Set when a lifted row would only lift with its ends pulled in to the chains we * loaded, so the numbers below describe the visible part and not the whole item. */ boolean clipped = FALSE; char *namedChainTrack = chainTrack; /* as the type line spells it, for the message below */ if (isBig && !isLifted) { chainTrack = netChainTrackName(tdb, chainTrack); chainTdb = netChainTdb(chainTrack); } if (otherOrg == NULL) { /* use first word in short track label */ otherOrg = firstWordInLine(cloneString(tdb->shortLabel)); } if (isBig) { net = bigNetLoadOne(tdb, seqName, start, sqlUnsigned(item), &clipped); if (net == NULL) errAbort("Couldn't find %s:%d at level %s in %s", seqName, start, item, tdb->track); } else { if (!hFindSplitTable(database, seqName, tdb->table, table, sizeof table, &hasBin)) errAbort("genericNetClick track %s not found", tdb->table); sqlSafef(query, sizeof(query), "select * from %s where tName = '%s' and tStart <= %d and tEnd > %d " "and level = %s", table, seqName, start, start, item); sr = sqlGetResult(conn, query); if ((row = sqlNextRow(sr)) == NULL) errAbort("Couldn't find %s:%d in %s", seqName, start, table); net = netAlignLoad(row+hasBin); sqlFreeResult(&sr); } tSize = net->tEnd - net->tStart; qSize = net->qEnd - net->qStart; if ((net->chainId != 0) && (!isBig || (chainTdb != NULL))) { netWinSize = min(winEnd-winStart, net->tEnd - net->tStart); printf("
\n"); /* Show alignment if the database exists and */ /* if there is a chromInfo table for that database and the sequence */ /* file exists. This means that alignments can be shown on the archive */ /* server (or in other cases) if there is a database with a chromInfo */ /* table, the sequences are available and there is an entry added to */ /* dbDb for the otherDb. */ if (chromSeqFileExists(otherDb, net->qName)) { if (netWinSize < 1000000) { int ns = max(winStart, net->tStart); int ne = min(winEnd, net->tEnd); if (ns < ne) { char id[20]; snprintf(id, sizeof(id), "%d", net->chainId); hgcAnchorWindow("htcChainAli", id, ns, ne, chainTrack, seqName); printf("View alignment details of parts of net within browser window.
\n"); } else { printf("Odd, net not in window
\n"); } } else { printf("To see alignment details zoom so that the browser window covers 1,000,000 bases or less.
\n"); } } if (isBig) { char idBuf[32]; safef(idBuf, sizeof idBuf, "%u", net->chainId); chain = chainLoadItemInRange(chainTdb, idBuf); } else chain = chainDbLoad(conn, database, chainTrack, seqName, net->chainId); if (chain != NULL) { /* print link to browser for otherDb only if otherDb is active */ if (hDbIsActive(otherDb)) chainToOtherBrowser(chain, otherDb, otherOrgBrowser, NULL); chainFree(&chain); } htmlHorizontalLine(); } else if (net->chainId != 0) { /* Only an isBig track with no chain track to follow gets here. */ if (isLifted) { char *sourceDb = trackDbSetting(tdb, "quickLiftDb"); htmlPrintf("
This net was lifted from %s, so its chains are not on this assembly " "and the alignment cannot be shown here.
\n", isEmpty(sourceDb) ? "another assembly" : sourceDb); } else { /* A hub whose type line names a chain track this assembly does not have. Say so * rather than dropping the whole section without a word. */ htmlPrintf("
This track's type line names the chain track %s, which is not on " "this assembly, so the alignment cannot be shown here.
\n", emptyForNull(namedChainTrack)); } htmlHorizontalLine(); } htmlPrintf("Type: %s
\n", net->type); printf("Level: %d
\n", (net->level+1)/2); htmlPrintf("%s position: %s:%d-%d
\n", org, net->tName, net->tStart+1, net->tEnd); htmlPrintf("%s position: %s:%d-%d
\n", otherOrg, net->qName, net->qStart+1, net->qEnd); printf("Strand: %c
\n", net->strand[0]); printLabeledNumber(NULL, "Score", net->score); if (net->chainId) { printf("Chain ID: %u
\n", net->chainId); printLabeledNumber(NULL, "Bases aligning", net->ali); if (net->qOver >= 0) printLabeledNumber(otherOrg, "parent overlap", net->qOver); if (net->qFar >= 0) printLabeledNumber(otherOrg, "parent distance", net->qFar); if (net->qDup >= 0) printLabeledNumber(otherOrg, "bases duplicated", net->qDup); } if (net->tN >= 0) printLabeledPercent(org, "N's", net->tN, tSize); if (net->qN >= 0) printLabeledPercent(otherOrg, "N's", net->qN, qSize); if (net->tTrf >= 0) printLabeledPercent(org, "tandem repeat (trf) bases", net->tTrf, tSize); if (net->qTrf >= 0) printLabeledPercent(otherOrg, "tandem repeat (trf) bases", net->qTrf, qSize); if (net->tR >= 0) printLabeledPercent(org, "RepeatMasker bases", net->tR, tSize); if (net->qR >= 0) printLabeledPercent(otherOrg, "RepeatMasker bases", net->qR, qSize); if (net->tOldR >= 0) printLabeledPercent(org, "old repeat bases", net->tOldR, tSize); if (net->qOldR >= 0) printLabeledPercent(otherOrg, "old repeat bases", net->qOldR, qSize); if (net->tNewR >= 0) printLabeledPercent(org, "new repeat bases", net->tOldR, tSize); if (net->qNewR >= 0) printLabeledPercent(otherOrg, "new repeat bases", net->qOldR, qSize); if (net->tEnd >= net->tStart) printLabeledNumber(org, "size", net->tEnd - net->tStart); if (net->qEnd >= net->qStart) printLabeledNumber(otherOrg, "size", net->qEnd - net->qStart); if (clipped) printf("
Fields above refer to the part of this chain or gap that could be " "placed on this assembly, not to the whole of it.
\n"); else printf("
Fields above refer to entire chain or gap, not just the part inside the window.
\n"); netAlignFree(&net); } void tfbsConsSites(struct trackDb *tdb, char *item) /* detail page for tfbsConsSites track */ { boolean printedPlus = FALSE; boolean printedMinus = FALSE; int start = cartInt(cart, "o"); struct sqlConnection *conn = hAllocConn(database); char table[HDB_MAX_TABLE_STRING]; boolean hasBin; char query[512]; struct sqlResult *sr; char **row; struct tfbsConsSites *tfbsConsSites; struct tfbsConsSites *tfbsConsSitesList = NULL; struct tfbsConsFactors *tfbsConsFactor; struct tfbsConsFactors *tfbsConsFactorList = NULL; boolean firstTime = TRUE; char *mappedId = NULL; genericHeader(tdb, item); if (!hFindSplitTable(database, seqName, tdb->table, table, sizeof table, &hasBin)) errAbort("tfbsConsSites track %s not found", tdb->table); sqlSafef(query, sizeof query, "select * from %s where name = '%s' and chrom = '%s' and chromStart = %d", table, item, seqName, start); sr = sqlGetResult(conn, query); while ((row = sqlNextRow(sr)) != NULL) { tfbsConsSites = tfbsConsSitesLoad(row+hasBin); slAddHead(&tfbsConsSitesList, tfbsConsSites); } sqlFreeResult(&sr); slReverse(&tfbsConsSitesList); if (!hFindSplitTable(database, seqName, "tfbsConsFactors", table, sizeof table, &hasBin)) errAbort("tfbsConsSites table tfbsConsFactors not found"); sqlSafef(query, sizeof query, "select * from %s where name = '%s' ", table, item); sr = sqlGetResult(conn, query); while ((row = sqlNextRow(sr)) != NULL) { tfbsConsFactor = tfbsConsFactorsLoad(row+hasBin); slAddHead(&tfbsConsFactorList, tfbsConsFactor); } sqlFreeResult(&sr); slReverse(&tfbsConsFactorList); if (tfbsConsFactorList) mappedId = cloneString(tfbsConsFactorList->ac); printf("Transcription Factor Binding Site information:
"); for(tfbsConsSites=tfbsConsSitesList ; tfbsConsSites != NULL ; tfbsConsSites = tfbsConsSites->next) { /* print each strand only once */ if ((printedMinus && (tfbsConsSites->strand[0] == '-')) || (printedPlus && (tfbsConsSites->strand[0] == '+'))) continue; if (!firstTime) htmlHorizontalLine(); else firstTime = FALSE; printf("Item: %s
\n", tfbsConsSites->name); if (mappedId != NULL) printCustomUrl(tdb, mappedId, FALSE); printf("Score: %d
\n", tfbsConsSites->score ); printf("zScore: %.2f
\n", tfbsConsSites->zScore ); printf("Strand: %s
\n", tfbsConsSites->strand); printPos(tfbsConsSites->chrom, tfbsConsSites->chromStart, tfbsConsSites->chromEnd, NULL, TRUE, tfbsConsSites->name); printedPlus = printedPlus || (tfbsConsSites->strand[0] == '+'); printedMinus = printedMinus || (tfbsConsSites->strand[0] == '-'); } if (tfbsConsFactorList) { htmlHorizontalLine(); printf("Transcription Factors known to bind to this site:
"); for(tfbsConsFactor =tfbsConsFactorList ; tfbsConsFactor != NULL ; tfbsConsFactor = tfbsConsFactor ->next) { if (!sameString(tfbsConsFactor->species, "N")) { printf("
Factor: %s
\n", tfbsConsFactor->factor); printf("Species: %s
\n", tfbsConsFactor->species); printf("SwissProt ID: %s
\n", sameString(tfbsConsFactor->id, "N")? "unknown": tfbsConsFactor->id); } } } printTrackHtml(tdb); hFreeConn(&conn); } void tfbsCons(struct trackDb *tdb, char *item) /* detail page for tfbsCons track */ { boolean printFactors = FALSE; boolean printedPlus = FALSE; boolean printedMinus = FALSE; int start = cartInt(cart, "o"); struct sqlConnection *conn = hAllocConn(database); char table[HDB_MAX_TABLE_STRING]; boolean hasBin; char query[512]; struct sqlResult *sr; char **row; struct tfbsCons *tfbs; struct tfbsCons *tfbsConsList = NULL; struct tfbsConsMap tfbsConsMap; boolean firstTime = TRUE; char *mappedId = NULL; genericHeader(tdb, item); if (!hFindSplitTable(database, seqName, tdb->table, table, sizeof table, &hasBin)) errAbort("tfbsCons track %s not found", tdb->table); sqlSafef(query, sizeof query, "select * from %s where name = '%s' and chrom = '%s' and chromStart = %d", table, item, seqName, start); sr = sqlGetResult(conn, query); while ((row = sqlNextRow(sr)) != NULL) { tfbs = tfbsConsLoad(row+hasBin); slAddHead(&tfbsConsList, tfbs); } sqlFreeResult(&sr); slReverse(&tfbsConsList); if (hTableExists(database, "tfbsConsMap")) { sqlSafef(query, sizeof query, "select * from tfbsConsMap where id = '%s'", tfbsConsList->name); sr = sqlGetResult(conn, query); if ((row = sqlNextRow(sr)) != NULL) { tfbsConsMapStaticLoad(row, &tfbsConsMap); mappedId = cloneString(tfbsConsMap.ac); } } sqlFreeResult(&sr); printf("Transcription Factor Binding Site information:
"); for(tfbs=tfbsConsList ; tfbs != NULL ; tfbs = tfbs->next) { if (!sameString(tfbs->species, "N")) printFactors = TRUE; /* print each strand only once */ if ((printedMinus && (tfbs->strand[0] == '-')) || (printedPlus && (tfbs->strand[0] == '+'))) continue; if (!firstTime) htmlHorizontalLine(); else firstTime = FALSE; printf("Item: %s
\n", tfbs->name); if (mappedId != NULL) printCustomUrl(tdb, mappedId, FALSE); printf("Score: %d
\n", tfbs->score ); printf("Strand: %s
\n", tfbs->strand); printPos(tfbsConsList->chrom, tfbs->chromStart, tfbs->chromEnd, NULL, TRUE, tfbs->name); printedPlus = printedPlus || (tfbs->strand[0] == '+'); printedMinus = printedMinus || (tfbs->strand[0] == '-'); } if (printFactors) { htmlHorizontalLine(); printf("Transcription Factors known to bind to this site:
"); for(tfbs=tfbsConsList ; tfbs != NULL ; tfbs = tfbs->next) { /* print only the positive strand when factors are on both strands */ if ((tfbs->strand[0] == '-') && printedPlus) continue; if (!sameString(tfbs->species, "N")) { printf("
Factor: %s
\n", tfbs->factor); printf("Species: %s
\n", tfbs->species); printf("SwissProt ID: %s
\n", sameString(tfbs->id, "N")? "unknown": tfbs->id); } } } printTrackHtml(tdb); hFreeConn(&conn); } void firstEF(struct trackDb *tdb, char *item) { int start = cartInt(cart, "o"); struct sqlConnection *conn = hAllocConn(database); char table[HDB_MAX_TABLE_STRING]; boolean hasBin; struct bed *bed; char query[512]; struct sqlResult *sr; char **row; boolean firstTime = TRUE; /* itemForUrl = item; */ genericHeader(tdb, item); printCustomUrl(tdb, item, FALSE); /* printCustomUrl(tdb, itemForUrl, item == itemForUrl); */ if (!hFindSplitTable(database, seqName, tdb->table, table, sizeof table, &hasBin)) errAbort("firstEF track %s not found", tdb->table); sqlSafef(query, sizeof query, "select * from %s where name = '%s' and chrom = '%s' and chromStart = %d", table, item, seqName, start); sr = sqlGetResult(conn, query); while ((row = sqlNextRow(sr)) != NULL) { if (firstTime) firstTime = FALSE; else htmlHorizontalLine(); bed = bedLoadN(row+hasBin, 6); printf("Item: %s
\n", bed->name); printf("Probability: %g
\n", bed->score / 1000.0); printf("Strand: %s
\n", bed->strand); printPos(bed->chrom, bed->chromStart, bed->chromEnd, NULL, TRUE, bed->name); } printTrackHtml(tdb); hFreeConn(&conn); } void doBed5FloatScore(struct trackDb *tdb, char *item) /* Handle click in BED 5+ track: BED 5 with 0-1000 score (for useScore * shading in hgTracks) plus real score for display in details page. */ { struct sqlConnection *conn = hAllocConn(database); struct sqlResult *sr = NULL; char table[HDB_MAX_TABLE_STRING]; boolean hasBin; struct bed5FloatScore *b5; struct dyString *query = dyStringNew(512); char **row; boolean firstTime = TRUE; int start = cartInt(cart, "o"); int bedSize = 5; if (!hFindSplitTable(database, seqName, tdb->table, table, sizeof table, &hasBin)) errAbort("doBed5FloatScore track %s not found", tdb->table); sqlDyStringPrintf(query, "select * from %s where chrom = '%s' and ", table, seqName); hAddBinToQuery(winStart, winEnd, query); sqlDyStringPrintf(query, "name = '%s' and chromStart = %d", item, start); sr = sqlGetResult(conn, query->string); while ((row = sqlNextRow(sr)) != NULL) { if (firstTime) firstTime = FALSE; else htmlHorizontalLine(); b5 = bed5FloatScoreLoad(row+hasBin); bedPrintPos((struct bed *)b5, 4, tdb); printf("Score: %f
\n", b5->floatScore); if (sameString(tdb->type, "bed5FloatScoreWithFdr")) { if (row[7] != NULL) printf("False Discovery Rate (FDR): %s%%
\n", row[7]); } } getBedTopScorers(conn, tdb, table, item, start, bedSize); sqlFreeResult(&sr); hFreeConn(&conn); /* printTrackHtml is done in genericClickHandlerPlus. */ } void doBed6FloatScore(struct trackDb *tdb, char *item) /* Handle click in BED 4+ track that's like BED 6 but with floating pt score */ { struct sqlConnection *conn = hAllocConn(database); struct sqlResult *sr = NULL; char table[HDB_MAX_TABLE_STRING]; boolean hasBin; struct bed6FloatScore *b6 = NULL; struct dyString *query = dyStringNew(512); char **row; boolean firstTime = TRUE; int start = cartInt(cart, "o"); genericHeader(tdb, item); if (!hFindSplitTable(database, seqName, tdb->table, table, sizeof table, &hasBin)) errAbort("doBed6FloatScore track %s not found", tdb->table); sqlDyStringPrintf(query, "select * from %s where chrom = '%s' and ", table, seqName); hAddBinToQuery(winStart, winEnd, query); sqlDyStringPrintf(query, "name = '%s' and chromStart = %d", item, start); sr = sqlGetResult(conn, query->string); while ((row = sqlNextRow(sr)) != NULL) { if (firstTime) firstTime = FALSE; else htmlHorizontalLine(); b6 = bed6FloatScoreLoad(row+hasBin); bedPrintPos((struct bed *)b6, 4, tdb); printf("Score: %f
\n", b6->score); printf("Strand: %s
\n", b6->strand); } sqlFreeResult(&sr); // Support for motif display if configured in trackDb // TODO - share code with factorSource char *motifPwmTable = trackDbSetting(tdb, "motifPwmTable"); struct dnaMotif *motif = NULL; if (motifPwmTable != NULL && sqlTableExists(conn, motifPwmTable)) { motif = loadDnaMotif(b6->name, motifPwmTable); if (motif == NULL) return; struct dnaSeq *seq = hDnaFromSeq(database, b6->chrom, b6->chromStart, b6->chromEnd, dnaLower); motifLogoAndMatrix(&seq, 1, motif); } hFreeConn(&conn); /* printTrackHtml is done in genericClickHandlerPlus. */ } void doColoredExon(struct trackDb *tdb, char *item) /* Print information for coloredExon type tracks. */ { struct sqlConnection *conn = hAllocConn(database); struct sqlResult *sr; char query[256]; char **row; genericHeader(tdb, item); sqlSafef(query, sizeof(query), "select chrom,chromStart,chromEnd,name,score,strand from %s where name='%s'", tdb->table, item); sr = sqlGetResult(conn, query); if ((row = sqlNextRow(sr)) != NULL) { struct bed *itemBed = bedLoad6(row); bedPrintPos(itemBed, 6, tdb); bedFree(&itemBed); } else { hPrintf("Could not find info for %s
\n", item); } sqlFreeResult(&sr); printTrackHtml(tdb); hFreeConn(&conn); } void doImageItemBed(struct trackDb *tdb, char *item) /* Print bed plus an image */ { } void doChromGraph(struct trackDb *tdb) /* Print information for coloredExon type tracks. */ { genericHeader(tdb, NULL); printTrackHtml(tdb); } static void genericContainerClick(struct sqlConnection *conn, char *containerType, struct trackDb *tdb, char *item, char *itemForUrl) /* Print page for container of some sort. */ { if (sameString(containerType, "multiWig")) { errAbort("It's suprising that multiWig container gets to hgc. It should go to hgTrackUi."); } else { errAbort("Unrecognized container type %s for %s", containerType, tdb->track); } } static void doBedMethyl(struct trackDb *tdb, char *item) /* Handle a click on a bedMethyl custom track */ { int start = cartInt(cart, "o"); struct sqlConnection *conn = hAllocConnTrack(CUSTOM_TRASH, tdb); database=CUSTOM_TRASH; tdb->table = trackDbSetting(tdb, "dbTableName"); genericBedClick(conn, tdb, item, start, 9); } static void doLongTabix(struct trackDb *tdb, char *item) /* Handle a click on a long range interaction */ { char *bigDataUrl = hashFindVal(tdb->settingsHash, "bigDataUrl"); struct bedTabixFile *btf = bedTabixFileMayOpen(bigDataUrl, NULL, 0, 0); char *chromName = cartString(cart, "c"); struct bed *list = bedTabixReadBeds(btf, chromName, winStart, winEnd, bedLoad5); bedTabixFileClose(&btf); unsigned maxWidth; struct longRange *longRangeList = parseLongTabix(list, &maxWidth, 0); struct longRange *longRange, *ourLongRange = NULL; unsigned itemNum = sqlUnsigned(item); unsigned count = slCount(longRangeList); double *doubleArray; AllocArray(doubleArray, count); int ii = 0; for(longRange = longRangeList; longRange; longRange = longRange->next, ii++) { if (longRange->id == itemNum) { ourLongRange = longRange; } doubleArray[ii] = longRange->score; } if (ourLongRange == NULL) errAbort("cannot find long range item with id %d\n", itemNum); printf("Item you clicked on:
\n"); printf(" ID: %u
\n", ourLongRange->id); if (!ourLongRange->hasColor) // if there's color, then there's no score in this format printf("Score: %g
\n", ourLongRange->score); unsigned padding = (ourLongRange->e - ourLongRange->s) / 10; int s = ourLongRange->s - padding; int e = ourLongRange->e + padding; if (s < 0 ) s = 0; int chromSize = hChromSize(database, seqName); if (e > chromSize) e = chromSize; char sStartPosBuf[1024], sEndPosBuf[1024]; char eStartPosBuf[1024], eEndPosBuf[1024]; // FIXME: longRange should store region starts, not centers sprintLongWithCommas(sStartPosBuf, ourLongRange->s - ourLongRange->sw/2 + 1); sprintLongWithCommas(sEndPosBuf, ourLongRange->s + ourLongRange->sw/2 + 1); sprintLongWithCommas(eStartPosBuf, ourLongRange->e - ourLongRange->ew/2 + 1); sprintLongWithCommas(eEndPosBuf, ourLongRange->e + ourLongRange->ew/2 + 1); char sWidthBuf[1024], eWidthBuf[1024]; char regionWidthBuf[1024]; sprintLongWithCommas(sWidthBuf, ourLongRange->sw); sprintLongWithCommas(eWidthBuf, ourLongRange->ew); sprintLongWithCommas(regionWidthBuf, ourLongRange->ew + ourLongRange->e - ourLongRange->s); if (differentString(ourLongRange->sChrom, ourLongRange->eChrom)) { printf("Current region: "); printf("%s:%s-%s (%s bp)
\n", ourLongRange->sChrom, sStartPosBuf, sEndPosBuf, ourLongRange->sChrom, sStartPosBuf,sEndPosBuf, sWidthBuf); printf("Paired region: "); printf("%s:%s-%s (%s bp)
\n", ourLongRange->eChrom, eStartPosBuf, eEndPosBuf, ourLongRange->eChrom, eStartPosBuf, eEndPosBuf, eWidthBuf); } else { printf("Lower region: "); printf("%s:%s-%s (%s bp)
\n", ourLongRange->sChrom, sStartPosBuf,sEndPosBuf, ourLongRange->sChrom, sStartPosBuf,sEndPosBuf, sWidthBuf); printf("Upper region: "); printf("%s:%s-%s (%s bp)
\n", ourLongRange->eChrom, eStartPosBuf, eEndPosBuf, ourLongRange->eChrom, eStartPosBuf, eEndPosBuf, eWidthBuf); printf("Interaction region: "); printf("%s:%s-%s (%s bp)
\n", ourLongRange->eChrom, sStartPosBuf, eEndPosBuf, ourLongRange->eChrom, sStartPosBuf, eEndPosBuf, regionWidthBuf); } if (ourLongRange->hasColor) return; struct aveStats *as = aveStatsCalc(doubleArray, count); printf("
Statistics on the scores of all items in window (go to track controls to set minimum score to display):\n"); printf("
| Q1 | %f |
| median | %f |
| Q3 | %f |
| average | %f |
| min | %f |
| max | %f |
| count | %d |
| total | %f |
| standard deviation | %f |
", trackHubSkipHubName(database), ourPos); int qs,qe; qChainRangePlusStrand(subChain, &qs, &qe); safef(position, 128, "%s:%d-%d", subChain->qName, qs-1, qe); otherPos = cloneString(addCommasToPos(otherDb, position)); htmlPrintf("%s position: ", trackHubSkipHubName(otherDb)); linkToOtherBrowser(otherDb, subChain->qName, qs-1, qe); htmlPrintf("%s
", otherPos); chainWinSize = min(winEnd-winStart, chain->tEnd - chain->tStart); if (otherTbf != NULL || (!startsWith("big", tdb->type) && sqlDatabaseExists(otherDb) && chromSeqFileExists(otherDb, chain->qName))) { if (chainWinSize < 1000000) { printf("View "); hgcAnchorSomewhere("htcChainAli", item, tdb->track, chain->tName); printf("DNA sequence alignment of whole window.
"); } else { printf("Zoom so that browser window covers 1,000,000 bases or less " "and return here to see alignment details.
\n"); } } char *liftDb = cloneString(trackDbSetting(tdb, "quickLiftDb")); char *quickLiftFile = cloneString(trackDbSetting(tdb, "quickLiftUrl")); struct quickLiftRegions *hr, *regions = quickLiftGetRegions(database, liftDb, quickLiftFile, chromName, seqStart, seqEnd); unsigned deletions = 0; unsigned insertions = 0; unsigned mismatches = 0; unsigned doubles = 0; for(hr = regions; hr; hr = hr->next) { switch(hr->type) { case QUICKTYPE_INSERT: insertions++; break; case QUICKTYPE_DEL: deletions++; break; case QUICKTYPE_DOUBLE: doubles++; break; case QUICKTYPE_MISMATCH: mismatches++; break; } } // If the click came from a specific difference (not one of the "identical" // regions), the matching row is shown in bold in the tables below. if (clickStart >= 0) printf("
The item you clicked on is shown in bold in the tables below.
\n"); if (deletions) { printf("
Deletions in Window:
"); printf("
| %s Position | %s Position | Bases | Alignment |
| %s | %s | ", ourPos, otherPos); printf("%.*s | ", hr->otherBaseCount, hr->otherBases); hgcAnchorSomewhereExt("htcChainAli", item, tdb->track, chain->tName, hr->chromStart - 10, hr->chromEnd + 10, tdb->track); printf("alignment |
Insertions in Window:
"); printf("
| %s Position | %s Position | Bases | Alignment |
| %s | %s | ", ourPos, otherPos); printf("%.*s | ", hr->baseCount, hr->bases); hgcAnchorSomewhereExt("htcChainAli", item, tdb->track, chain->tName, hr->chromStart - 10, hr->chromEnd + 10, tdb->track); printf("alignment |
Double Gaps in Window:
"); printf("
| %s Position | %s Position | # Bases in %s | #Bases in %s | Alignment |
| %s | %s | ", ourPos, otherPos); printf("%d | %d | ", hr->baseCount, hr->otherBaseCount); hgcAnchorSomewhereExt("htcChainAli", item, tdb->track, chain->tName, hr->chromStart - 10, hr->chromEnd + 10, tdb->track); printf("alignment |
Mismatches in Window:
"); printf("
| %s Position | %s Position | Change | Alignment |
| %s | %s | ", ourPos, otherPos); printf("%.*s -> %.*s | ", hr->otherBaseCount, hr->otherBases, hr->baseCount, hr->bases); hgcAnchorSomewhereExt("htcChainAli", item, tdb->track, chain->tName, hr->chromStart - 10, hr->chromEnd + 10, tdb->track); printf("alignment |
\n", bothWords[0], item, bothWords[1]); shouldBeTwo = chopLine(bigImagePath, bothWords); if (shouldBeTwo != 2) errAbort("bigItemImagePath setting for %s track incorrect. Needs to be \"itemImagePath
\n", bothWords[0], item, bothWords[1]); } if ((sameString(tdb->table,"altLocations") || sameString(tdb->table,"fixLocations")) && strchr(item,'_')) { // Truncate item (alt/fix sequence name) at colon if found: char itemCpy[strlen(item)+1]; safecpy(itemCpy, sizeof(itemCpy), item); char *p = strchr(itemCpy, ':'); if (p) *p = '\0'; char *hgsid = cartSessionId(cart); char *desc = sameString(tdb->table, "altLocations") ? "alternate haplotype" : "fix patch"; printf("" "Show this %s placed on its chromosome
\n", hgsid, itemCpy, desc); } printTrackHtml(tdb); freez(&dupe); hFreeConn(&conn); } void genericClickHandler(struct trackDb *tdb, char *item, char *itemForUrl) /* Put up generic track info */ { genericClickHandlerPlus(tdb, item, itemForUrl, NULL); } void savePosInTextBox(char *chrom, int start, int end) /* Save basic position/database info in text box and hidden var. Positions becomes chrom:start-end*/ { char position[128]; char *newPos; snprintf(position, 128, "%s:%d-%d", chrom, start, end); newPos = addCommasToPos(database, position); cgiMakeTextVar("getDnaPos", newPos, strlen(newPos) + 2); cgiContinueHiddenVar("db"); } char *hgTablesUrl(boolean usePos, char *track) /* Make up URL for table browser. */ { struct dyString *url = dyStringNew(0); dyStringAppend(url, "../cgi-bin/hgTables?"); dyStringAppend(url, cartSidUrlString(cart)); dyStringPrintf(url, "&db=%s", database); if (usePos) { dyStringPrintf(url, "&position=%s:%d-%d", seqName, winStart+1, winEnd); dyStringAppend(url, "&hgta_regionType=range"); } if (track != NULL) { struct trackDb *tdb = hashFindVal(trackHash, track); if (tdb != NULL) { char *grp = tdb->grp; if (grp != NULL && grp[0] != 0) { dyStringPrintf(url, "&hgta_group=%s", grp); dyStringPrintf(url, "&hgta_track=%s", track); dyStringPrintf(url, "&hgta_table=%s", track); } } } return dyStringCannibalize(&url); } char *traceUrl(char *traceId) /* Make up URL for trace archive. */ { struct dyString *url = dyStringNew(0); dyStringAppend(url, "https://www.ncbi.nlm.nih.gov/Traces/trace.cgi?"); dyStringPrintf(url, "cmd=retrieve&size=1&val=%s&", traceId); dyStringAppend(url, "file=trace&dopt=trace"); return dyStringCannibalize(&url); } void doGetDna1() /* Do first get DNA dialog. */ { struct hTableInfo *hti = NULL; char *tbl = cgiUsualString("table", ""); if (dbIsFound && tbl[0] != 0) { char rootName[HDB_MAX_TABLE_STRING]; char parsedChrom[HDB_MAX_CHROM_STRING]; hParseTableName(database, tbl, rootName, parsedChrom); if (!trackHubDatabase(database)) hti = hFindTableInfo(database, seqName, rootName); } char *thisOrg = hOrganism(database); cartWebStart(cart, database, "Get DNA in Window (%s/%s)", trackHubSkipHubName(database), trackHubSkipHubName(thisOrg)); printf("
Get DNA for
\n"); printf(""); if (dbIsFound) puts("Note: The \"Mask repeats\" option applies only to \"Get DNA\", not to \"Extended case/color options\".
"); } boolean dnaIgnoreTrack(char *track) /* Return TRUE if this is one of the tracks too boring * to put DNA on. */ { return (sameString("cytoBand", track) || sameString("gcPercent", track) || sameString("gold", track) || sameString("gap", track) || startsWith("mouseSyn", track)); } struct customTrack *getCtList() /* initialize theCtList if necessary and return it */ { if (theCtList == NULL) theCtList = customTracksParseCart(database, cart, NULL, NULL); return(theCtList); } struct trackDb *tdbForCustomTracks() /* Load custom tracks (if any) and translate to list of trackDbs */ { struct customTrack *ctList = getCtList(); struct customTrack *ct; struct trackDb *tdbList = NULL, *tdb; for (ct=ctList; ct != NULL; ct=ct->next) { AllocVar(tdb); tdb->track = ct->tdb->track; tdb->table = ct->tdb->table; tdb->shortLabel = ct->tdb->shortLabel; tdb->type = ct->tdb->type; tdb->longLabel = ct->tdb->longLabel; tdb->visibility = ct->tdb->visibility; tdb->priority = ct->tdb->priority; tdb->colorR = ct->tdb->colorR; tdb->colorG = ct->tdb->colorG; tdb->colorB = ct->tdb->colorB; tdb->altColorR = ct->tdb->altColorR; tdb->altColorG = ct->tdb->altColorG; tdb->altColorB = ct->tdb->altColorB; tdb->useScore = ct->tdb->useScore; tdb->private = ct->tdb->private; tdb->url = ct->tdb->url; tdb->grp = ct->tdb->grp; tdb->canPack = ct->tdb->canPack; trackDbPolish(tdb); slAddHead(&tdbList, tdb); } slReverse(&tdbList); return(tdbList); } static struct trackDb *getCustomTrackTdb(char *table) /* Find the trackDb structure for a custom track table. */ { struct customTrack *ctList = getCtList(); struct customTrack *ct = NULL; for (ct = ctList; ct != NULL; ct = ct->next) if (sameString(table, ct->tdb->track)) return ct->tdb; return NULL; } struct trackDb *getTdbForTrackName(char *trackName) /* Given a track name (which may have ct_ or hub_ prepended), tdb */ { assert(trackHash != NULL); if (isCustomTrack(trackName)) return getCustomTrackTdb(trackName); else if (isHubTrack(trackName)) return hubConnectAddHubForTrackAndFindTdb( database, trackName, NULL, trackHash); else return hashMustFindVal(trackHash, trackName); } struct customTrack *lookupCt(char *name) /* Return custom track for name, or NULL. */ { struct customTrack *ct; for (ct=getCtList(); ct != NULL; ct=ct->next) if (sameString(name, ct->tdb->track)) return(ct); return(NULL); } void parseSs(char *ss, char **retPslName, char **retFaName, char **retQName) /* Parse space separated 'ss' item. */ { static char buf[512*2]; int wordCount; char *words[4]; /* SECURITY (refs #38054): ss comes from the cart variable of the same name, or from * the item name, and a visitor controls both. A legitimate value is a short triple * of trash file paths, so anything that does not fit is a bug or an attack. safecpy * aborts instead of writing past the end of the buffer, which matches the errAborts * below on the other malformed cases. */ safecpy(buf, sizeof buf, ss); wordCount = chopLine(buf, words); if (wordCount < 1) errAbort("Empty user cart variable ss."); *retPslName = words[0]; if (retFaName != NULL) { if (wordCount < 2) errAbort("Expecting psl filename and fa filename in cart variable ss, but only got one word: %s", ss); *retFaName = words[1]; } if (retQName != NULL) { if (wordCount < 3) errAbort("Expecting psl filename, fa filename and query name in cart variable ss, but got this: %s", ss); *retQName = words[2]; } } boolean ssFilesExist(char *ss) /* Return TRUE if both files in ss exist. -- Copied from hgTracks! */ { char *faFileName, *pslFileName; parseSs(ss, &pslFileName, &faFileName, NULL); return fileExists(pslFileName) && fileExists(faFileName); } struct trackDb *tdbForUserPsl() /* Load up user's BLAT results into trackDb. */ { char *ss = cartOptionalString(cart, "ss"); if ((ss != NULL) && !ssFilesExist(ss)) { ss = NULL; cartRemove(cart, "ss"); } if (ss == NULL) return(NULL); else { struct trackDb *tdb; AllocVar(tdb); tdb->track = cloneString(USER_PSL_TRACK_NAME); tdb->table = cloneString(USER_PSL_TRACK_NAME); tdb->shortLabel = cloneString(USER_PSL_TRACK_LABEL); tdb->type = cloneString("psl"); tdb->longLabel = cloneString(USER_PSL_TRACK_LONGLABEL); tdb->visibility = tvFull; tdb->priority = 11.0; trackDbPolish(tdb); return(tdb); } } struct trackDb *rFindUnderstandableTrack(char *db, struct trackDb *tdb) // If any leaf is usable in getting DNA then that leaf's tdb is returned. { if (tdb->subtracks != NULL) return rFindUnderstandableTrack(db,tdb->subtracks); if (fbUnderstandTrack(db, tdb) && !dnaIgnoreTrack(tdb->table)) return tdb; else return NULL; } boolean forestHasUnderstandableTrack(char *db, struct trackDb *tdb) // TRUE if any leaf is usable in getting DNA. { return (rFindUnderstandableTrack(db, tdb) != NULL); } struct trackDb* loadTracks() /* load native tracks, cts, userPsl and track Hubs and return tdbList */ { struct trackDb *tdbList = hTrackDb(database); struct trackDb *ctdbList = tdbForCustomTracks(); struct trackDb *utdbList = tdbForUserPsl(); struct grp *pGrpList = NULL; struct trackDb *hubList = hubCollectTracks(database, &pGrpList); ctdbList = slCat(ctdbList, tdbList); ctdbList = slCat(ctdbList, hubList); tdbList = slCat(utdbList, ctdbList); return tdbList; } static boolean parseDnaPos(char *db, char *pos, char **retChrom, int *retStart, int *retEnd) /* Parse a position string that may be chrom:start-end or just chrom:pos (single base). */ { if (hgParseChromRange(db, pos, retChrom, retStart, retEnd)) return TRUE; /* check for single position like chrX:67774500 (colon but no dash) */ char *colon = strchr(pos, ':'); if (colon != NULL && strchr(colon, '-') == NULL) { struct dyString *dy = dyStringCreate("%s-%s", pos, colon + 1); boolean result = hgParseChromRange(db, dy->string, retChrom, retStart, retEnd); dyStringFree(&dy); return result; } return FALSE; } void doGetDnaExtended1() /* Do extended case/color get DNA options. */ { boolean revComp = cartUsualBoolean(cart, "hgSeq.revComp", FALSE); boolean maskRep = cartUsualBoolean(cart, "hgSeq.maskRepeats", FALSE); int padding5 = cartUsualInt(cart, "hgSeq.padding5", 0); int padding3 = cartUsualInt(cart, "hgSeq.padding3", 0); int lineWidth = cartUsualInt(cart, "lineWidth", 60); char *casing = cartUsualString(cart, "hgSeq.casing", ""); char *repMasking = cartUsualString(cart, "hgSeq.repMasking", ""); boolean caseUpper= FALSE; char *pos = NULL; struct trackDb *tdbList = loadTracks(); cartWebStart(cart, database, "Extended DNA Case/Color"); if (NULL != (pos = stripCommas(cartOptionalString(cart, "getDnaPos")))) parseDnaPos(database, pos, &seqName, &winStart, &winEnd); if (winEnd - winStart > 5000000) { printf("Please zoom in to 5 million bases or less to color the DNA"); return; } printf("
Extended DNA Case/Color Options
\n"); puts( "Use this page to highlight features in genomic DNA text. " "DNA covered by a particular track can be highlighted by " "case, underline, bold, italic, or color. See below for " "details about color, and for examples. Tracks in " ""hide" display mode are not shown in the grid below."); if (cgiBooleanDefined("hgSeq.maskRepeats")) cartSetBoolean(cart, "hgSeq.maskRepeats", maskRep); if (*repMasking != 0) cartSetString(cart, "hgSeq.repMasking", repMasking); if (maskRep) { struct trackDb *rtdb; char *visString = cartOptionalString(cart, "rmsk"); for (rtdb = tdbList; rtdb != NULL; rtdb=rtdb->next) { if (startsWith(rtdb->table, "rmsk")) break; } printf("
Note: repeat masking style from previous page will not apply to this page.\n"); if ((rtdb != NULL) && ((visString == NULL) || !sameString(visString, "hide"))) printf("Use the case/color options for the RepeatMasker track below.
\n"); else printf("Unhide the RepeatMasker track in the genome browser, then return to this page and use the case/color options for the RepeatMasker track below.
\n"); } cartSetInt(cart, "padding5", padding5); cartSetInt(cart, "padding3", padding3); if (sameString(casing, "upper")) caseUpper = TRUE; if (*casing != 0) cartSetString(cart, "hgSeq.casing", casing); printf("
\n"); printf("Coloring Information and Examples
\n"); puts("The color values range from 0 (darkest) to 255 (lightest) and are additive.\n"); puts("The examples below show a few ways to highlight individual tracks, " "and their interplay. It's good to keep it simple at first. It's easy " "to make pretty, but completely cryptic, displays with this feature."); puts( "- "
"
- To put exons from RefSeq Genes in upper case red text, check the " "appropriate box in the Toggle Case column and set the color to pure " "red, RGB (255,0,0). Upon submitting, any RefSeq Gene within the " "designated chromosomal interval will now appear in red capital letters.\n" "
- To see the overlap between RefSeq Genes and Genscan predictions try " "setting the RefSeq Genes to red (255,0,0) and Genscan to green (0,255,0). " "Places where the RefSeq Genes and Genscan overlap will be painted yellow " "(255,255,0).\n" "
- To get a level-of-coverage effect for tracks like Spliced Ests with " "multiple overlapping items, initially select a darker color such as deep " "green, RGB (0,64,0). Nucleotides covered by a single EST will appear dark " "green, while regions covered with more ESTs get progressively brighter — " "saturating at 4 ESTs." "
- Another track can be used to mask unwanted features. Setting the " "RepeatMasker track to RGB (255,255,255) will white-out Genscan predictions " "of LINEs but not mainstream host genes; masking with RefSeq Genes will show " "what is new in the gene prediction sector." "
Further Details and Ideas
"); puts("Copying and pasting the web page output to a text editor such as Word " "will retain upper case but lose colors and other formatting. That is still " "useful because other web tools such as " "NCBI Blast " "can be set to ignore lower case. To fully capture formatting such as color " "and underlining, view the output as \"source\" in your web browser, or download " "it, or copy the output page into an html editor.
"); puts("The default line width of 60 characters is standard, but if you have " "a reasonable sized monitor it's useful to set this higher - to 125 characters " "or more. You can see more DNA at once this way, and fewer line breaks help " "in finding DNA strings using the web browser search function.
"); puts("Be careful about requesting complex formatting for a very large " "chromosomal region. After all the html tags are added to the output page, " "the file size may exceed size limits that your browser, clipboard, and " "other software can safely display. The tool will format 10 Mb and more, however.
"); trackDbFreeList(&tdbList); } void doGetBlastPep(char *readName, char *table) /* get predicted protein */ { int qStart; struct psl *psl; int start, end; struct sqlResult *sr; struct sqlConnection *conn = hAllocConn(database); struct dnaSeq *tSeq; char query[256], **row; char fullTable[HDB_MAX_TABLE_STRING]; boolean hasBin; char *buffer, *str; int i, j; char *ptr; start = cartInt(cart, "o"); if (!hFindSplitTable(database, seqName, table, fullTable, sizeof fullTable, &hasBin)) errAbort("doGetBlastPep track %s not found", table); sqlSafef(query, sizeof query, "select * from %s where qName = '%s' and tName = '%s' and tStart=%d", fullTable, readName, seqName, start); sr = sqlGetResult(conn, query); if ((row = sqlNextRow(sr)) == NULL) errAbort("Couldn't find alignment for %s at %d", readName, start); psl = pslLoad(row+hasBin); sqlFreeResult(&sr); hFreeConn(&conn); printf("");
end = psl->tEnd;
if (psl->strand[1] == '+')
end = psl->tStarts[psl->blockCount - 1] + psl->blockSizes[psl->blockCount - 1] *3;
if ((ptr = strchr(readName, '.')) != NULL)
*ptr++ = 0;
printf(">%s-%s\n", readName,database);
tSeq = hDnaFromSeq(database, psl->tName, start, end, dnaLower);
if (psl->strand[1] == '-')
{
start = psl->tSize - end;
reverseComplement(tSeq->dna, tSeq->size);
}
str = buffer = needMem(psl->qSize + 1);
qStart = 0;
for (i=0; iblockCount; ++i)
{
int ts = psl->tStarts[i] - start;
int sz = psl->blockSizes[i];
for (;qStart < psl->qStarts[i]; qStart++)
*str++ = 'X';
for (j=0; jdna[codonStart];
if ((*str = lookupCodon(codon)) == 0)
*str = '*';
str++;
qStart++;
}
}
*str = 0;
printLines(stdout, buffer, 50);
printf(" ");
}
void doGetDna2()
/* Do second DNA dialog (or just fetch DNA) */
{
char *tbl = cgiUsualString("table", "");
char *action = cgiUsualString("submit", "");
int itemCount;
char *pos = NULL;
char *chrom = NULL;
int start = 0;
int end = 0;
if (sameString(action, EXTENDED_DNA_BUTTON))
{
doGetDnaExtended1();
return;
}
// This output probably should be just text/plain but
// trying to support the fancy warn handler box requires html.
// But we want to keep it very simple and close to a plain text dump.
cartHtmlStart("DNA");
puts("");
if (tbl[0] == 0)
{
itemCount = 1;
if ( NULL != (pos = stripCommas(cartOptionalString(cart, "getDnaPos"))) &&
parseDnaPos((dbIsFound ? database : NULL), pos, &chrom, &start, &end))
{
hgSeqRange(database, chrom, start, end, '?', "dna");
}
else
{
hgSeqRange(database, seqName, cartInt(cart, "l"), cartInt(cart, "r"),
'?', "dna");
}
}
else
{
struct hTableInfo *hti = NULL;
char rootName[HDB_MAX_TABLE_STRING];
char parsedChrom[HDB_MAX_CHROM_STRING];
/* use the values from the dnaPos dialog box */
if (!( NULL != (pos = stripCommas(cartOptionalString(cart, "getDnaPos"))) &&
parseDnaPos(database, pos, &chrom, &start, &end)))
{
/* if can't get DnaPos from dialog box, use "o" and "t" */
start = cartInt(cart, "o");
end = cartInt(cart, "t");
}
else
{
seqName = chrom;
}
/* Table might be a custom track if it's not in the database,
* or bigBed if it is in the database but has only one column called 'fileName';
* in which case, just get DNA as if no table were given. */
hParseTableName(database, tbl, rootName, parsedChrom);
if (!trackHubDatabase(database))
hti = hFindTableInfo(database, seqName, rootName);
if (hti == NULL || hti->startField[0] == 0)
{
itemCount = 1;
hgSeqRange(database, seqName, start, end, '?', tbl);
}
else
{
char *where = NULL;
char *item = cgiUsualString("i", "");
char buf[256];
if ((hti->nameField[0] != 0) && (item[0] != 0))
{
sqlSafef(buf, sizeof(buf), "%s = '%s'", hti->nameField, item);
where = buf;
}
itemCount = hgSeqItemsInRange(database, tbl, seqName, start, end, where);
}
}
if (itemCount == 0)
printf("\n# No results returned from query.\n\n");
puts("");
}
struct hTableInfo *ctToHti(struct customTrack *ct)
/* Create an hTableInfo from a customTrack. */
{
struct hTableInfo *hti;
AllocVar(hti);
hti->rootName = cloneString(ct->tdb->table);
hti->isPos = TRUE;
hti->isSplit = FALSE;
hti->hasBin = FALSE;
hti->type = cloneString(ct->tdb->type);
int fieldCount = 3;
if (sameWord(ct->dbTrackType, "bedDetail"))
fieldCount = ct->fieldCount - 2;
else if (sameWord(ct->dbTrackType, "pgSnp"))
fieldCount = 4;
else
fieldCount = ct->fieldCount;
if (fieldCount >= 3)
{
strncpy(hti->chromField, "chrom", 32);
strncpy(hti->startField, "chromStart", 32);
strncpy(hti->endField, "chromEnd", 32);
}
if (fieldCount >= 4)
{
strncpy(hti->nameField, "name", 32);
}
if (fieldCount >= 5)
{
strncpy(hti->scoreField, "score", 32);
}
if (fieldCount >= 6)
{
strncpy(hti->strandField, "strand", 32);
}
if (fieldCount >= 8)
{
strncpy(hti->cdsStartField, "thickStart", 32);
strncpy(hti->cdsEndField, "thickEnd", 32);
hti->hasCDS = TRUE;
}
if (fieldCount >= 12)
{
strncpy(hti->countField, "blockCount", 32);
strncpy(hti->startsField, "chromStarts", 32);
strncpy(hti->endsSizesField, "blockSizes", 32);
hti->hasBlocks = TRUE;
}
return(hti);
}
struct hTableInfo *htiForUserPsl()
/* Create an hTableInfo for user's BLAT results. */
{
struct hTableInfo *hti;
AllocVar(hti);
hti->rootName = cloneString(USER_PSL_TRACK_NAME);
hti->isPos = TRUE;
hti->isSplit = FALSE;
hti->hasBin = FALSE;
hti->type = cloneString("psl");
strncpy(hti->chromField, "tName", 32);
strncpy(hti->startField, "tStart", 32);
strncpy(hti->endField, "tEnd", 32);
strncpy(hti->nameField, "qName", 32);
/* psl can be scored... but strictly speaking, does not have a score field! */
strncpy(hti->strandField, "strand", 32);
hti->hasCDS = FALSE;
strncpy(hti->countField, "blockCount", 32);
strncpy(hti->startsField, "tStarts", 32);
strncpy(hti->endsSizesField, "tSizes", 32);
hti->hasBlocks = TRUE;
return(hti);
}
struct bed *bedFromUserPsl()
/* Load up user's BLAT results into bedList. */
{
struct bed *bedList = NULL;
char *ss = cartOptionalString(cart, "ss");
if ((ss != NULL) && ! ssFilesExist(ss))
{
ss = NULL;
cartRemove(cart, "ss");
}
if (ss == NULL)
return(NULL);
else
{
struct lineFile *f;
struct psl *psl;
enum gfType qt, tt;
char *faFileName, *pslFileName;
int i;
parseSs(ss, &pslFileName, &faFileName, NULL);
pslxFileOpen(pslFileName, &qt, &tt, &f);
while ((psl = pslNext(f)) != NULL)
{
struct bed *bed;
AllocVar(bed);
bed->chrom = cloneString(seqName);
bed->chromStart = psl->tStart;
bed->chromEnd = psl->tEnd;
bed->name = cloneString(psl->qName);
bed->score = pslScore(psl);
if ((psl->strand[0] == '-' && psl->strand[1] == '+') ||
(psl->strand[0] == '+' && psl->strand[1] == '-'))
strncpy(bed->strand, "-", 2);
else
strncpy(bed->strand, "+", 2);
bed->thickStart = bed->chromStart;
bed->thickEnd = bed->chromEnd;
bed->blockCount = psl->blockCount;
bed->chromStarts = needMem(bed->blockCount * sizeof(int));
bed->blockSizes = needMem(bed->blockCount * sizeof(int));
for (i=0; i < bed->blockCount; i++)
{
bed->chromStarts[i] = psl->tStarts[i];
bed->blockSizes[i] = psl->blockSizes[i];
}
if (qt == gftProt)
for (i=0; i < bed->blockCount; i++)
{
/* If query is protein, blockSizes are in aa units; fix 'em. */
bed->blockSizes[i] *= 3;
}
if (psl->strand[1] == '-')
{
/* psl: if target strand is '-', flip the coords.
* (this is the target part of pslRc from src/lib/psl.c) */
for (i=0; i < bed->blockCount; ++i)
{
bed->chromStarts[i] =
psl->tSize - (bed->chromStarts[i] +
bed->blockSizes[i]);
}
reverseInts(bed->chromStarts, bed->blockCount);
reverseInts(bed->blockSizes, bed->blockCount);
assert(bed->chromStart == bed->chromStarts[0]);
}
/* translate absolute starts to relative starts (after handling
* target-strand coord-flipping) */
for (i=0; i < bed->blockCount; i++)
{
bed->chromStarts[i] -= bed->chromStart;
}
slAddHead(&bedList, bed);
pslFree(&psl);
}
lineFileClose(&f);
slReverse(&bedList);
return(bedList);
}
}
void addColorToRange(int r, int g, int b, struct rgbColor *colors, int start, int end)
/* Add rgb values to colors array from start to end. Don't let values
* exceed 255 */
{
struct rgbColor *c;
int rr, gg, bb;
int i;
for (i=start; i");
printf(">%s:%d-%d %s\n", seqName, winStart+1, winEnd,
(isRc ? "(reverse complement)" : ""));
seq = hDnaFromSeq(database, seqName, winStart, winEnd, dnaLower);
if (isRc)
reverseComplement(seq->dna, seq->size);
if (defaultUpper)
touppers(seq->dna);
AllocArray(colors, winSize);
struct trackDb* tdb;
for (tdb = tdbList; tdb != NULL; tdb = tdb->next)
{
char *track = tdb->track;
char *table = tdb->table;
struct featureBits *fbList = NULL, *fb;
struct customTrack *ct = lookupCt(track);
if (sameString(USER_PSL_TRACK_NAME, table)
|| ct != NULL
|| ( tdbVisLimitedByAncestors(cart,tdb,TRUE,TRUE) != tvHide
&& forestHasUnderstandableTrack(database, tdb) ) )
{
char buf[256];
int r,g,b;
/* to save a LOT of time, don't fetch track features unless some
* coloring/formatting has been specified for them. */
boolean hasSettings = FALSE;
safef(buf, sizeof(buf), "%s_u", track);
hasSettings |= cgiBoolean(buf);
safef(buf, sizeof(buf), "%s_b", track);
hasSettings |= cgiBoolean(buf);
safef(buf, sizeof(buf), "%s_i", track);
hasSettings |= cgiBoolean(buf);
safef(buf, sizeof(buf), "%s_case", track);
hasSettings |= cgiBoolean(buf);
safef(buf, sizeof(buf), "%s_red", track);
hasSettings |= (cgiOptionalInt(buf, 0) != 0);
safef(buf, sizeof(buf), "%s_green", track);
hasSettings |= (cgiOptionalInt(buf, 0) != 0);
safef(buf, sizeof(buf), "%s_blue", track);
hasSettings |= (cgiOptionalInt(buf, 0) != 0);
if (! hasSettings)
continue;
if (sameString(USER_PSL_TRACK_NAME, track))
{
struct hTableInfo *hti = htiForUserPsl();
struct bedFilter *bf;
struct bed *bedList, *bedList2;
AllocVar(bf);
bedList = bedFromUserPsl();
bedList2 = bedFilterListInRange(bedList, bf, seqName, winStart,
winEnd);
fbList = fbFromBed(database, track, hti, bedList2, winStart, winEnd,
TRUE, FALSE);
bedFreeList(&bedList);
bedFreeList(&bedList2);
}
else if (ct != NULL)
{
struct hTableInfo *hti = ctToHti(ct);
struct bedFilter *bf;
struct bed *bedList2, *ctBedList = NULL;
AllocVar(bf);
if (ct->dbTrack && (!sameString(tdb->type, "vcf")))
{
struct bed *bed;
int fieldCount = ct->fieldCount;
if ((ct->dbTrackType != NULL) && sameString(ct->dbTrackType, "pgSnp"))
fieldCount = 4;
char query[512];
int rowOffset;
char **row;
struct sqlConnection *conn = hAllocConn(CUSTOM_TRASH);
struct sqlResult *sr = NULL;
sqlSafef(query, sizeof(query), "select * from %s", ct->dbTableName);
sr = hRangeQuery(conn, ct->dbTableName, seqName,
winStart, winEnd, NULL, &rowOffset);
while ((row = sqlNextRow(sr)) != NULL)
{
bed = bedLoadN(row+rowOffset, fieldCount);
if (bf == NULL || bedFilterOne(bf, bed))
{
struct bed *copy = cloneBed(bed);
slAddHead(&ctBedList, copy);
}
}
sqlFreeResult(&sr);
hFreeConn(&conn);
}
else if (sameString(ct->dbTrackType, "bigBed"))
{
struct lm *lm = lmInit(0);
struct bigBedInterval *bb, *bbList = bigBedIntervalQuery(ct->bbiFile, seqName, winStart, winEnd, 0, lm);
char *bedRow[32];
char startBuf[16], endBuf[16];
for (bb = bbList; bb != NULL; bb = bb->next)
{
bigBedIntervalToRow(bb, seqName, startBuf, endBuf, bedRow, ArraySize(bedRow));
struct bed *bed = bedLoadN(bedRow, ct->bbiFile->definedFieldCount);
slAddHead(&ctBedList, bed);
}
}
else
{
ctBedList = ct->bedList;
}
if (startsWith("vcf", tdb->type))
{
fbList = vcfLoadInterval(ct->tdb, winStart, winEnd);
}
else
{
bedList2 = bedFilterListInRange(ctBedList, bf, seqName, winStart,
winEnd);
fbList = fbFromBed(database, track, hti, bedList2, winStart, winEnd,
TRUE, FALSE);
bedFreeList(&bedList2);
}
if (!ct->bedList)
bedFreeList(&ctBedList);
}
else
{
if (tdb->subtracks)
{
struct slRef *refLeaves = trackDbListGetRefsToDescendantLeaves(tdb->subtracks);
struct slRef *refLeaf = NULL;
while ((refLeaf = slPopHead(&refLeaves)) != NULL)
{
struct trackDb *tdbLeaf = refLeaf->val;
if (tdbVisLimitedByAncestors(cart,tdbLeaf,TRUE,TRUE) != tvHide
&& fbUnderstandTrack(database, tdbLeaf)
&& !dnaIgnoreTrack(tdbLeaf->table))
{
struct featureBits *fbLeafList;
if (startsWith("vcf", tdbLeaf->type))
fbLeafList = vcfLoadInterval(tdbLeaf, winStart, winEnd);
else if (startsWith("big", tdbLeaf->type))
fbLeafList = getBigBedFbList(tdbLeaf, seqName, winStart, winEnd);
else
fbLeafList = fbGetRange(database, tdbLeaf->table, seqName, winStart, winEnd);
if (fbLeafList != NULL)
fbList = slCat(fbList,fbLeafList);
}
freeMem(refLeaf);
}
}
else
{
if (startsWith("vcf", tdb->type))
fbList = vcfLoadInterval(tdb, winStart, winEnd);
else if (startsWith("big", tdb->type))
fbList = getBigBedFbList(tdb, seqName, winStart, winEnd);
else
fbList = fbGetRange(database, tdb->table, seqName, winStart, winEnd);
}
}
/* Flip underline/italic/bold bits. */
getDnaHandleBits(track, "u", uBits, winStart, winEnd, isRc, fbList);
getDnaHandleBits(track, "b", bBits, winStart, winEnd, isRc, fbList);
getDnaHandleBits(track, "i", iBits, winStart, winEnd, isRc, fbList);
/* Toggle case if necessary. */
safef(buf, sizeof buf, "%s_case", track);
if (cgiBoolean(buf))
{
for (fb = fbList; fb != NULL; fb = fb->next)
{
DNA *dna;
int start = fb->start - winStart;
int end = fb->end - winStart;
int size = fb->end - fb->start;
if (isRc)
reverseIntRange(&start, &end, seq->size);
dna = seq->dna + start;
if (defaultUpper)
toLowerN(dna, size);
else
toUpperN(dna, size);
}
}
/* Add in RGB values if necessary. */
safef(buf, sizeof buf, "%s_red", track);
r = cartInt(cart, buf);
safef(buf, sizeof buf, "%s_green", track);
g = cartInt(cart, buf);
safef(buf, sizeof buf, "%s_blue", track);
b = cartInt(cart, buf);
if (r != 0 || g != 0 || b != 0)
{
for (fb = fbList; fb != NULL; fb = fb->next)
{
int s = fb->start - winStart;
int e = fb->end - winStart;
if (isRc)
reverseIntRange(&s, &e, winEnd - winStart);
addColorToRange(r, g, b, colors, s, e);
}
}
}
}
cfm = cfmNew(0, lineWidth, FALSE, FALSE, stdout, 0);
for (i=0; isize; ++i)
{
struct rgbColor *color = colors+i;
int c = (color->r<<16) + (color->g<<8) + color->b;
cfmOutExt(cfm, seq->dna[i], c,
bitReadOne(uBits, i), bitReadOne(bBits, i), bitReadOne(iBits, i));
}
cfmFree(&cfm);
freeDnaSeq(&seq);
bitFree(&uBits);
bitFree(&iBits);
bitFree(&bBits);
}
void medlineLinkedTermLine(char *title, char *text, char *search, char *keyword)
/* Produce something that shows up on the browser as
* TITLE: value
* with the value hyperlinked to medline using a specified search term. */
{
char *encoded = cgiEncode(search);
char *encodedKeyword = cgiEncode(keyword);
printf("%s: ", title);
if (sameWord(text, "n/a") || sameWord(text, "none"))
printf("n/a
\n");
else
{
printf("%s
\n", text);
}
freeMem(encoded);
}
void medlineLinkedLine(char *title, char *text, char *search)
/* Produce something that shows up on the browser as
* TITLE: value
* with the value hyperlinked to medline. */
{
char *encoded = cgiEncode(search);
printf("%s: ", title);
if (sameWord(text, "n/a"))
printf("n/a
\n");
else
{
printf("%s
\n", text);
}
freeMem(encoded);
}
void medlineProductLinkedLine(char *title, char *text)
/* Produce something that shows up on the browser as
* TITLE: value
* with the value hyperlinked to medline.
* Replaces commas in the product name with spaces, as commas sometimes
* interfere with PubMed search */
{
subChar(text, ',', ' ');
medlineLinkedLine(title, text, text);
}
void appendAuthor(struct dyString *dy, char *gbAuthor, int len)
/* Convert from Kent,W.J. to Kent WJ and append to dy.
* gbAuthor gets eaten in the process.
* Also strip web URLs since Entrez doesn't like those. */
{
char buf[MAXURLSIZE];
char *ptr;
if (len >= sizeof(buf))
warn("author %s too long to process", gbAuthor);
else
{
memcpy(buf, gbAuthor, len);
buf[len] = 0;
stripChar(buf, '.');
subChar(buf, ',' , ' ');
if ((ptr = strstr(buf, " http://")) != NULL)
*ptr = 0;
if ((ptr = strstr(buf, " https://")) != NULL)
*ptr = 0;
dyStringAppend(dy, buf);
dyStringAppend(dy, " ");
}
}
void gbToEntrezAuthor(char *authors, struct dyString *dy)
/* Convert from Genbank author format:
* Kent,W.J., Haussler,D. and Zahler,A.M.
* to Entrez search format:
* Kent WJ,Haussler D,Zahler AM
*/
{
char *s = authors, *e;
/* Parse first authors, which will be terminated by '.,' */
while ((e = strstr(s, ".,i ")) != NULL)
{
int len = e - s + 1;
appendAuthor(dy, s, len);
s += len+2;
}
if ((e = strstr(s, " and")) != NULL)
{
int len = e - s;
appendAuthor(dy, s, len);
s += len+4;
}
if ((s = skipLeadingSpaces(s)) != NULL && s[0] != 0)
{
int len = strlen(s);
appendAuthor(dy, s, len);
}
}
/* --- !!! Riken code is under development Fan. 4/16/02 */
void printRikenInfo(char *acc, struct sqlConnection *conn )
/* Print Riken annotation info */
{
struct sqlResult *sr;
char **row;
char query[512];
char *seqid, *accession, *comment;
// char *qualifier, *anntext, *datasrc, *srckey, *href, *evidence;
accession = acc;
sqlSafef(query, sizeof(query),
"select seqid from rikenaltid where altid='%s';", accession);
sr = sqlMustGetResult(conn, query);
row = sqlNextRow(sr);
if (row != NULL)
{
seqid=cloneString(row[0]);
sqlSafef(query, sizeof(query),
"select Qualifier, Anntext, Datasrc, Srckey, Href, Evidence "
"from rikenann where seqid='%s';", seqid);
sqlFreeResult(&sr);
sr = sqlMustGetResult(conn, query);
row = sqlNextRow(sr);
while (row !=NULL)
{
// qualifier = row[0]; unused variable
// anntext = row[1]; unused variable
// datasrc = row[2]; unused variable
// srckey = row[3]; unused variable
// href = row[4]; unused variable
// evidence = row[5]; unused variable
row = sqlNextRow(sr);
}
sqlSafef(query, sizeof(query),
"select comment from rikenseq where id='%s';", seqid);
sqlFreeResult(&sr);
sr = sqlMustGetResult(conn, query);
row = sqlNextRow(sr);
if (row != NULL)
{
comment = row[0];
printf("Riken/comment: %s
\n",comment);
}
}
}
void printGeneCards(char *db, char *geneName)
/* Print out a link to GeneCards (Human only).
* Caller passes the assembly db so quickLifted callers can pass the source
* (quickLiftDb) rather than the destination hub_NNN_. */
{
if (startsWith("hg", db) && isNotEmpty(geneName))
{
printf("GeneCards: "
"%s
\n",
geneName, geneName);
}
}
int getImageId(struct sqlConnection *conn, char *acc)
/* get the image id for a clone, or 0 if none */
{
int imageId = 0;
if (sqlTableExists(conn, imageCloneTable))
{
struct sqlResult *sr;
char **row;
char query[128];
sqlSafef(query, sizeof(query),
"select imageId from %s where acc = '%s'",imageCloneTable, acc);
sr = sqlMustGetResult(conn, query);
row = sqlNextRow(sr);
if (row != NULL)
imageId = sqlUnsigned(row[0]);
sqlFreeResult(&sr);
}
return imageId;
}
void htcDisplayMrna(char *acc)
/* Display mRNA available from genback or seq table.. */
{
struct dnaSeq *seq = hGenBankGetMrna(database, acc, NULL);
if (seq == NULL)
errAbort("mRNA sequence %s not found", acc);
cartHtmlStart("mRNA sequence");
printf("");
faWriteNext(stdout, seq->name, seq->dna, seq->size);
printf("");
dnaSeqFree(&seq);
}
static int getEstTranscriptionDir(struct sqlConnection *conn, struct psl *psl)
/* get the direction of transcription for an EST; return splice support count */
{
char query[256], estOrient[64];
sqlSafef(query, sizeof(query),
"select intronOrientation from %s.estOrientInfo where chrom = '%s' and chromStart = %d and name = '%s'",
database, psl->tName, psl->tStart, psl->qName);
if (sqlQuickQuery(conn, query, estOrient, sizeof(estOrient)) != NULL)
return sqlSigned(estOrient) * ((psl->strand[0] == '+') ? 1 : -1);
else
return 0;
}
static struct gbWarn *checkGbWarn(struct sqlConnection *conn, char *acc)
/* check if there is a gbWarn entry for this accession, return NULL if none */
{
struct gbWarn *gbWarn = NULL;
if (sqlTableExists(conn, gbWarnTable))
gbWarn = sqlQueryObjs(conn, (sqlLoadFunc)gbWarnLoad, sqlQuerySingle,
"SELECT * FROM %s WHERE acc = \"%s\"", gbWarnTable, acc);
return gbWarn;
}
static void printGbWarn(char *acc, struct gbWarn *gbWarn)
/* print descriptive information about an accession in the gbWarn table */
{
char *msg = NULL;
switch (gbWarn->reason) {
case gbWarnInvitroNorm:
msg = "is from the InVitroGen/Genoscope full-length library. Some of the entries "
"associated with this dataset appear to have been aligned to the reference "
"genome and the sequences subsequently modified to match the genome. This "
"process may have resulted in apparent high-quality alignments to pseudogenes.";
break;
case gbWarnAthRage:
msg = "is from the Athersys RAGE library. These sequences were created by inducing expression and may not "
"be an indication of in vivo expression.";
break;
case gbWarnOrestes:
msg = "is from an ORESTES library. This protocol includes a PCR step subject to genomic contamination.";
break;
}
assert(msg != NULL);
char *msg2= "Care should be taken in using alignments of this sequence as evidence of transcription.";
printf("Warning: %s %s %s
\n", acc, msg, msg2);
}
static void printRnaSpecs(struct trackDb *tdb, char *acc, struct psl *psl)
/* Print auxiliarry info on RNA. */
{
struct dyString *dy = dyStringNew(1024);
struct sqlConnection *conn = hAllocConn(database);
struct sqlConnection *conn2= hAllocConn(database);
struct sqlResult *sr;
char **row;
char rgdEstId[512];
char query[256];
char *type,*direction,*orgFullName,*library,*clone,*sex,*tissue,
*development,*cell,*cds,*description, *author,*geneName,
*date,*productName;
// char *source; unused variable
// int seqSize,fileSize; unused variables
// long fileOffset; unused variable
// char *extFile; unused variable
boolean hasVersion = hHasField(database, gbCdnaInfoTable, "version");
boolean haveGbSeq = sqlTableExists(conn, gbSeqTable);
char *seqTbl = haveGbSeq ? gbSeqTable : "seq";
char *version = NULL;
struct trackDb *tdbRgdEst;
char *chrom = cartString(cart, "c");
int start = cartInt(cart, "o");
int end = cartUsualInt(cart, "t",0);
struct gbWarn *gbWarn = checkGbWarn(conn, acc);
/* This sort of query and having to keep things in sync between
* the first clause of the select, the from clause, the where
* clause, and the results in the row ... is really tedious.
* One of my main motivations for going to a more object
* based rather than pure relational approach in general,
* and writing 'autoSql' to help support this. However
* the pure relational approach wins for pure search speed,
* and these RNA fields are searched. So it looks like
* the code below stays. Be really careful when you modify
* it.
*
* Uses the gbSeq table if available, otherwise use seq for older databases.
*/
sqlDyStringPrintf(dy,
"select g.type,g.direction,"
"so.name,o.name,l.name,m.name,"
"se.name,t.name,dev.name,ce.name,cd.name,"
"des.name,a.name,gene.name,p.name,"
"gbS.size,g.moddate,gbS.gbExtFile,gbS.file_offset,gbS.file_size ");
/* If the gbCdnaInfoTAble table has a "version" column then will show it */
if (hasVersion)
{
sqlDyStringPrintf(dy,
", g.version ");
}
sqlDyStringPrintf(dy,
" from %s g,%s gbS,%s so,%s o,%s l,%s m,%s se,%s t,"
"%s dev,%s ce,%s cd,%s des,%s a,%s gene,%s p"
" where g.acc = '%s' and g.id = gbS.id ",
gbCdnaInfoTable,seqTbl, sourceTable, organismTable, libraryTable, mrnaCloneTable, sexTable, tissueTable, developmentTable, cellTable, cdsTable, descriptionTable, authorTable, geneNameTable, productNameTable, acc);
sqlDyStringPrintf(dy,
"and g.source = so.id and g.organism = o.id "
"and g.library = l.id and g.mrnaClone = m.id "
"and g.sex = se.id and g.tissue = t.id "
"and g.development = dev.id and g.cell = ce.id "
"and g.cds = cd.id and g.description = des.id "
"and g.author = a.id and g.geneName = gene.id "
"and g.productName = p.id");
sr = sqlMustGetResult(conn, dy->string);
row = sqlNextRow(sr);
if (row != NULL)
{
type=row[0];direction=row[1];
// source=row[2]; unused variable
orgFullName=row[3];library=row[4];clone=row[5];
sex=row[6];tissue=row[7];development=row[8];cell=row[9];cds=row[10];description=row[11];
author=row[12];geneName=row[13];productName=row[14];
// seqSize = sqlUnsigned(row[15]); unused variable
date = row[16];
// ext_file = row[17]; unused variable
// fileOffset=sqlUnsigned(row[18]); unused variable
// fileSize=sqlUnsigned(row[19]); unused variable
boolean isEst = sameWord(type, "est");
if (hasVersion)
{
version = row[20];
}
/* Now we have all the info out of the database and into nicely named
* local variables. There's still a few hoops to jump through to
* format this prettily on the web with hyperlinks to NCBI. */
printf("Information on %s %s
\n", acc);
printf("Description: %s
\n", description);
if (gbWarn != NULL)
printGbWarn(acc, gbWarn);
medlineLinkedLine("Gene", geneName, geneName);
medlineProductLinkedLine("Product", productName);
dyStringClear(dy);
gbToEntrezAuthor(author, dy);
medlineLinkedLine("Author", author, dy->string);
printf("Organism: ");
printf("",
cgiEncode(orgFullName));
printf("%s
\n", orgFullName);
printf("Tissue: %s
\n", tissue);
printf("Development stage: %s
\n", development);
printf("Cell line: %s
\n", cell);
printf("Sex: %s
\n", sex);
printf("Library: %s
\n", library);
printf("Clone: %s
\n", clone);
if (isEst)
{
printf("Read direction: ");
if (direction[0] != '0')
printf("%s' (guessed from GenBank description)
\n", direction);
else
printf("unknown (can't guess from GenBank description)
");
}
else
printf("CDS: %s
\n", cds);
printf("Date: %s
\n", date);
if (hasVersion)
{
printf("Version: %s
\n", version);
}
/* print RGD EST Report link if it is Rat genome and it has a link to RGD */
if (sameWord(organism, "Rat"))
{
if (hTableExists(database, "rgdEstLink"))
{
sqlSafef(query, sizeof(query),
"select id from %s.rgdEstLink where name = '%s';", database, acc);
if (sqlQuickQuery(conn2, query, rgdEstId, sizeof(rgdEstId)) != NULL)
{
tdbRgdEst = hashFindVal(trackHash, "rgdEst");
printf("RGD EST Report: ");
printf("", tdbRgdEst->url, rgdEstId);
printf("RGD:%s
\n", rgdEstId);
}
}
}
if (isEst && hTableExists(database, "estOrientInfo") && (psl != NULL))
{
int estOrient = getEstTranscriptionDir(conn2, psl);
if (estOrient != 0)
printf("EST transcribed from %c strand (supported by %d splice sites).
\n",
(estOrient > 0 ? '+' : '-' ), abs(estOrient));
}
if (hGenBankHaveSeq(database, acc, NULL))
{
printf("%s sequence: ", type);
hgcAnchorSomewhere("htcDisplayMrna", acc, tdb->track, seqName);
printf("%s
\n", acc);
}
}
else
{
warn("Couldn't find %s in %s table", gbCdnaInfoTable, acc);
}
if (end != 0 && differentString(chrom,"0") && isNotEmpty(chrom))
{
printf("Position: "
"",
hgTracksPathAndSettings(), database, chrom, start+1, end);
printf("%s:%d-%d
\n", chrom, start+1, end);
}
gbWarnFree(&gbWarn);
sqlFreeResult(&sr);
dyStringFree(&dy);
hFreeConn(&conn);
hFreeConn(&conn2);
}
static char *aliTrackParam()
/* "&aliTrack=