a1cfd4eb6e2f1e0eb0e8756fd02ebcb602c77687 kent Tue Jun 28 13:08:30 2022 -0700 Making base-by-base alignment display work for bigChains and bigPsls. diff --git src/hg/hgc/hgc.c src/hg/hgc/hgc.c index 8efffb4..0c22794 100644 --- src/hg/hgc/hgc.c +++ src/hg/hgc/hgc.c @@ -257,30 +257,31 @@ #include "bPlusTree.h" #include "customFactory.h" #include "iupac.h" #include "clinvarSubLolly.h" #include "jsHelper.h" #include "errCatch.h" #include "htslib/bgzf.h" #include "htslib/kstring.h" #include "pipeline.h" #include "genark.h" #include "chromAlias.h" static char *rootDir = "hgcData"; #define LINESIZE 70 /* size of lines in comp seq feature */ +#define MAX_DISPLAY_QUERY_SEQ_SIZE 5000000 // Big enough for HLA alts struct cart *cart; /* User's settings. */ char *seqName; /* Name of sequence we're working on. */ int winStart, winEnd; /* Bounds of sequence. */ char *database; /* Name of mySQL database. */ char *organism; /* Colloquial name of organism. */ char *genome; /* common name, e.g. Mouse, Human */ char *scientificName; /* Scientific name of organism. */ /* for earlyBotCheck() function at the beginning of main() */ #define delayFraction 0.5 /* standard penalty is 1.0 for most CGIs */ /* this one is 0.5 */ boolean issueBotWarning = FALSE; struct hash *trackHash; /* A hash of all tracks - trackDb valued */ @@ -3224,39 +3225,71 @@ } void pslDumpHtml(struct psl *pslList) /* print out psl header and data */ { struct psl* psl; printf("<PRE><TT>\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("</TT></PRE>\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 = twoBitReadSeqFrag(otherTbf, psl->qName, 0, 0); + *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++; } boolean showEvery = sameString(item, "PrintAllSequences"); boolean showAll = trackDbSettingOn(tdb, "showAll"); unsigned seqTypeField = bbExtraFieldIndex(bbi, "seqType"); struct bigBedInterval *bb, *bbList = NULL; // If showAll is on, show all alignments with this qName, not just the // selected one. if (showEvery) @@ -3313,57 +3346,60 @@ char startBuf[16], endBuf[16]; int lastChromId = -1; char chromName[bbi->chromBpt->keySize+1]; boolean firstTime = TRUE; 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= pslFromBigPsl(chromName, bb, seqTypeField, &seq, &cdsStr); + struct psl *psl= getPslAndSeq(tdb, chromName, bb, seqTypeField, &seq, &cdsStr); 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 (differentString(item, "PrintAllSequences") && firstTime && !isEmpty(seq)) // if there is a query sequence { firstTime = FALSE; printf("<H3>Links to sequence:</H3>\n"); printf("<UL>\n"); if (!isEmpty(cdsStr)) // if we have CDS { puts("<LI>\n"); hgcAnchorSomewhere("htcTranslatedBigPsl", item, "translate", seqName); printf("Translated Amino Acid Sequence</A> from Query Sequence\n"); puts("</LI>\n"); } + if (psl->qSize <= MAX_DISPLAY_QUERY_SEQ_SIZE) + { puts("<LI>\n"); hgcAnchorSomewhere("htcBigPslSequence", item, tdb->track, seqName); printf("Query Sequence</A> \n"); puts("</LI>\n"); printf("</UL>\n"); } } } + } char *sort = cartUsualString(cart, "sort", pslSortList[0]); pslSortListByVar(&pslList, sort); if (showEvery) printf("<H3>Genomic Alignments</H3>"); else printf("<H3>%s/Genomic Alignments</H3>", item); if (showEvery || pslIsProtein(pslList)) printAlignmentsSimple(pslList, start, "htcBigPslAli", tdb->table, item); else printAlignmentsExtra(pslList, start, "htcBigPslAli", "htcBigPslAliInWindow", tdb->table, item); pslFreeList(&pslList); printItemDetailsHtml(tdb, item); @@ -3488,30 +3524,49 @@ { 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("<div class='readableWidth'>"); puts(html); puts("</div>"); } hPrintf("<BR>\n"); } +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 (startsWith("big", tdb->type)) + { + char *fileName = trackDbSetting(tdb, "bigDataUrl"); + char *linkFileName = trackDbSetting(tdb, "linkDataUrl"); + 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; @@ -3585,79 +3640,88 @@ 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); printf("Open %s browser</A> at position corresponding to the part of chain that is in this window.<BR>\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 genericChainClick(struct sqlConnection *conn, struct trackDb *tdb, char *item, int start, char *otherDb) /* Handle click in chain track, at least the basics. */ { +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)) { - char *fileName = bbiNameFromSettingOrTable(tdb, conn, tdb->table); - char *linkFileName = trackDbSetting(tdb, "linkDataUrl"); - chain = chainLoadIdRangeHub(database, fileName, linkFileName, seqName, winStart, winEnd, atoi(item)); - 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; } } } -else - { - chain = chainLoadIdRange(database, tdb->table, seqName, winStart, winEnd, atoi(item)); - } 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"); @@ -3670,31 +3734,31 @@ 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 = hChromSeq(otherDb, subChain->qName, qStart, qEnd); + 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); printf("<B>%s position:</B> <A HREF=\"%s?%s&db=%s&position=%s:%d-%d\">%s:%d-%d</A>" " size: %d <BR>\n", trackHubSkipHubName(thisOrg), hgTracksName(), cartSidUrlString(cart), database, chain->tName, chain->tStart+1, chain->tEnd, chain->tName, chain->tStart+1, chain->tEnd, chain->tEnd-chain->tStart); printf("<B>Strand:</B> %c<BR>\n", chain->qStrand); qChainRangePlusStrand(chain, &qs, &qe); @@ -3751,31 +3815,33 @@ } sqlFreeResult(&sr); printf("<BR>\n"); } printf("<BR>Fields above refer to entire chain or gap, not just the part inside the window.<BR>\n"); printf("<BR>\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 (!startsWith("big", tdb->type) && sqlDatabaseExists(otherDb) && chromSeqFileExists(otherDb, chain->qName)) +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</A> details of parts of chain within browser " "window.<BR>\n"); } else { printf("Zoom so that browser window covers 1,000,000 bases or less " "and return here to see alignment details.<BR>\n"); } } @@ -4708,34 +4774,30 @@ char *hgsid = cartSessionId(cart); char *desc = sameString(tdb->table, "altLocations") ? "alternate haplotype" : "fix patch"; printf("<A HREF=\"hgTracks?hgsid=%s&virtModeType=singleAltHaplo&singleAltHaploId=%s\">" "Show this %s placed on its chromosome</A><BR>\n", hgsid, itemCpy, desc); } printTrackHtml(tdb); freez(&dupe); hFreeConn(&conn); } void genericClickHandler(struct trackDb *tdb, char *item, char *itemForUrl) /* Put up generic track info */ { -#ifdef OLD /* Called now by cartWebStart... */ -jsIncludeFile("jquery.js", NULL); -jsIncludeFile("utils.js",NULL); -#endif /* OLD */ 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"); } @@ -4880,30 +4942,53 @@ 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) @@ -6317,60 +6402,64 @@ printf("BROWSER | SIZE IDENTITY SCAFFOLD STRAND START END QUERY START END TOTAL\n"); printf("-----------------------------------------------------------------------------------------------------\n"); for (isClicked = 1; isClicked >= 0; isClicked -= 1) { for (psl = pslList; psl != NULL; psl = psl->next) { if (isPslToPrintByClick(psl, startFirst, isClicked)) { char otherString[512]; char *qName = itemIn; if (sameString(itemIn, "PrintAllSequences")) qName = psl->qName; safef(otherString, sizeof(otherString), "%d&aliTable=%s", psl->tStart, tableName); printf("<A HREF=\"%s&db=%s&position=%s%%3A%d-%d\">browser</A> | ", hgTracksPathAndSettings(), database, psl->tName, psl->tStart+1, psl->tEnd); + if (psl->qSize <= MAX_DISPLAY_QUERY_SEQ_SIZE) // Only anchor if small enough hgcAnchorWindow(hgcCommand, qName, psl->tStart, psl->tEnd, otherString, psl->tName); - printf("%5d %5.1f%% %9s %s %9d %9d %20s %5d %5d %5d</A>", + printf("%5d %5.1f%% %9s %s %9d %9d %20s %5d %5d %5d", psl->match + psl->misMatch + psl->repMatch, 100.0 - pslCalcMilliBad(psl, TRUE) * 0.1, skipChr(psl->tName), psl->strand, psl->tStart + 1, psl->tEnd, psl->qName, psl->qStart+1, psl->qEnd, psl->qSize); + if (psl->qSize <= MAX_DISPLAY_QUERY_SEQ_SIZE) + printf("</A>"); printf("\n"); } } } printf("</TT></PRE>"); } void printAlignmentsExtra(struct psl *pslList, int startFirst, char *hgcCommand, char *hgcCommandInWindow, char *tableName, char *itemIn) /* Print list of mRNA alignments with special "in window" alignment function. */ { if (pslList == NULL || tableName == NULL) return; printAlignmentsSimple(pslList, startFirst, hgcCommand, tableName, itemIn); struct psl *psl = pslList; for (psl = pslList; psl != NULL; psl = psl->next) { if ( pslTrimToTargetRange(psl, winStart, winEnd) != NULL && !startsWith("xeno", tableName) && !(startsWith("user", tableName) && pslIsProtein(psl)) && psl->tStart == startFirst && sameString(psl->tName, seqName) + && psl->qSize <= MAX_DISPLAY_QUERY_SEQ_SIZE ) { char otherString[512]; safef(otherString, sizeof(otherString), "%d&aliTable=%s", psl->tStart, tableName); hgcAnchorSomewhere(hgcCommandInWindow, itemIn, otherString, psl->tName); printf("<BR>View details of parts of alignment within browser window</A>.<BR>\n"); } } } void printAlignments(struct psl *pslList, int startFirst, char *hgcCommand, char *tableName, char *itemIn) /* Print list of mRNA alignments. */ { @@ -7727,31 +7816,31 @@ 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, 12); if (sameString(bed->name, acc) && (bb->start == start) && (bb->end == end)) { bb->next = NULL; break; } } if (bb == NULL) errAbort("item %s not found in range %s:%d-%d in bigBed %s (%s)", acc, chrom, start, end, tdb->table, fileName); unsigned seqTypeField = bbExtraFieldIndex(bbi, "seqType"); -psl = pslFromBigPsl(seqName, bb, seqTypeField, &seq, &cdsString); +psl = getPslAndSeq(tdb, seqName, bb, seqTypeField, &seq, &cdsString); if (cdsString) genbankParseCds(cdsString, &cdsStart, &cdsEnd); if (seq == NULL) { printf("Sequence for %s not available.\n", psl->qName); return; } struct dnaSeq *rnaSeq = newDnaSeq(seq, strlen(seq), acc); enum gfType type = gftRna; if (pslIsProtein(psl)) type = gftProt; showSomeAlignment(psl, rnaSeq, type, 0, rnaSeq->size, NULL, cdsStart, cdsEnd); } @@ -7788,31 +7877,31 @@ struct bbiFile *bbi = bigBedFileOpenAlias(fileName, chromAliasFindAliases); struct bigBedInterval *bb, *bbList = bigBedIntervalQuery(bbi, chrom, start, end, 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, 12); if (sameString(bed->name, acc)) { bb->next = NULL; break; } } unsigned seqTypeField = bbExtraFieldIndex(bbi, "seqType"); -wholePsl = pslFromBigPsl(seqName, bb, seqTypeField, &seq, &cdsString); +wholePsl = getPslAndSeq(tdb, seqName, bb, seqTypeField, &seq, &cdsString); if (seq == NULL) { printf("Sequence for %s not available.\n", wholePsl->qName); return; } if (cdsString) genbankParseCds(cdsString, &cdsStart, &cdsEnd); if (wholePsl->tStart >= winStart && wholePsl->tEnd <= winEnd) partPsl = wholePsl; else partPsl = pslTrimToTargetRange(wholePsl, winStart, winEnd); struct dnaSeq *rnaSeq = newDnaSeq(seq, strlen(seq), acc); showSomePartialDnaAlignment(partPsl, wholePsl, rnaSeq, @@ -8061,109 +8150,118 @@ else partPsl = pslTrimToTargetRange(wholePsl, winStart, winEnd); if (startsWith("xeno", aliTable)) errAbort("htcCdnaAliInWindow does not support translated alignments."); else showSomePartialDnaAlignment(partPsl, wholePsl, rnaSeq, NULL, cdsStart, cdsEnd); } void htcChainAli(char *item) /* Draw detailed alignment representation of a chain. */ { struct chain *chain; struct psl *fatPsl, *psl = NULL; -int id = atoi(item); char *track = cartString(cart, "o"); -char *type = trackTypeInfo(track); +struct trackDb *tdb = getTdbForTrackName(track); +char *type = cloneString(tdb->type); char *typeWords[2]; char *otherDb = NULL, *org = NULL, *otherOrg = NULL; struct dnaSeq *qSeq = NULL; char name[128]; - /* Figure out other database. */ if (chopLine(type, typeWords) < ArraySize(typeWords)) errAbort("type line for %s is short in trackDb", track); otherDb = typeWords[1]; if (! sameWord(otherDb, "seq")) { otherOrg = hOrganism(otherDb); } org = hOrganism(database); /* Load up subset of chain and convert it to part of psl * that just fits inside of window. */ -chain = chainLoadIdRange(database, track, seqName, winStart, winEnd, id); +chain = chainLoadItemInRange(tdb, item); if (chain->blockList == NULL) { printf("None of chain is actually in the window"); return; } fatPsl = chainToPsl(chain); chainFree(&chain); psl = pslTrimToTargetRange(fatPsl, winStart, winEnd); pslFree(&fatPsl); +struct twoBitFile *otherTbf = getOtherTwoBitUrl(tdb); if (sameWord(otherDb, "seq")) { qSeq = hExtSeqPart(database, psl->qName, psl->qStart, psl->qEnd); safef(name, sizeof name, "%s", psl->qName); } -else +else if (otherDb != NULL) { qSeq = loadGenomePart(otherDb, psl->qName, psl->qStart, psl->qEnd); safef(name, sizeof name, "%s.%s", otherOrg, psl->qName); } +else if (otherTbf != NULL) + { + qSeq = twoBitReadSeqFragLower(otherTbf, psl->qName, psl->qStart, psl->qEnd); + safef(name, sizeof name, "%s", psl->qName); + } +if (qSeq == NULL) + { + errAbort("Can't find query sequence in htcChainAli"); + } char title[1024]; safef(title, sizeof title, "%s %s vs %s %s ", (otherOrg == NULL ? "" : otherOrg), psl->qName, org, psl->tName ); htmlFramesetStart(title); showSomeAlignment(psl, qSeq, gftDnaX, psl->qStart, psl->qEnd, name, 0, 0); } void htcChainTransAli(char *item) /* Draw detailed alignment representation of a chain with translated protein */ { struct chain *chain; struct psl *fatPsl, *psl = NULL; -int id = atoi(item); char *track = cartString(cart, "o"); char *type = trackTypeInfo(track); char *typeWords[2]; char *otherDb = NULL, *org = NULL, *otherOrg = NULL; struct dnaSeq *qSeq = NULL; char name[128]; int cdsStart = cgiInt("qs"); int cdsEnd = cgiInt("qe"); /* Figure out other database. */ if (chopLine(type, typeWords) < ArraySize(typeWords)) errAbort("type line for %s is short in trackDb", track); otherDb = typeWords[1]; if (! sameWord(otherDb, "seq")) { otherOrg = hOrganism(otherDb); } org = hOrganism(database); /* Load up subset of chain and convert it to part of psl * that just fits inside of window. */ -chain = chainLoadIdRange(database, track, seqName, winStart, winEnd, id); +struct trackDb *tdb = getTdbForTrackName(track); +chain = chainLoadItemInRange(tdb, item); if (chain->blockList == NULL) { printf("None of chain is actually in the window"); return; } fatPsl = chainToPsl(chain); chainFree(&chain); psl = pslTrimToTargetRange(fatPsl, winStart, winEnd); pslFree(&fatPsl); if (sameWord(otherDb, "seq")) { qSeq = hExtSeq(database, psl->qName); @@ -8877,41 +8975,30 @@ struct bbiFile *bbi = bigBedFileOpenAlias(fileName, chromAliasFindAliases); struct lm *lm = lmInit(0); struct bigBedInterval *bb, *bbList = bigBedIntervalQuery(bbi, seqName, winStart, winEnd, 0, lm); struct genePred *gpList = NULL; for (bb = bbList; bb != NULL; bb = bb->next) { struct genePred *gp = (struct genePred *)genePredFromBigGenePred(seqName, bb); if (sameString(gp->name, geneName)) slAddHead(&gpList, gp); } lmCleanup(&lm); return gpList; } -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; -} - static struct genePred *getGenePredForPosition(char *table, char *geneName) /* Build a genePred list for the given table and gene name. */ { struct genePred *gpList = NULL; if (isCustomTrack(table)) { struct trackDb *tdb = getCustomTrackTdb(table); gpList = getGenePredForPositionBigGene(tdb, geneName); } else if (isHubTrack(table)) { struct trackDb *tdb = hubConnectAddHubForTrackAndFindTdb( database, table, NULL, trackHash); gpList = getGenePredForPositionBigGene(tdb, geneName); } @@ -8996,31 +9083,31 @@ int lastChromId = -1; char chromName[bbi->chromBpt->keySize+1]; 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); // 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 sequence if (sameString(bedRow[3], acc)) { char *cdsStr, *seq; - pslFromBigPsl(chromName, bb, seqTypeField, &seq, &cdsStr); + getPslAndSeq(tdb, chromName, bb, seqTypeField, &seq, &cdsStr); if (translate) { struct genbankCds cds; if (!genbankCdsParse(cdsStr, &cds)) errAbort("can't parse CDS for %s: %s", acc, cdsStr); int protBufSize = ((cds.end-cds.start)/3)+4; char *prot = needMem(protBufSize); seq[cds.end] = '\0'; dnaTranslateSome(seq+cds.start, prot, protBufSize); cartHtmlStart("Protein Translation of query sequence"); displayProteinPrediction(acc, prot); return; @@ -14738,30 +14825,31 @@ && strcmp(otherName,"pig") && strcmp(otherName,"cat") && strcmp(otherName,"dog") && strcmp(otherName,"mouse") && strcmp(otherName,"rat") && strcmp(otherName,"chicken") && strcmp(otherName,"fugu") && strcmp(otherName,"tetra") && strcmp(otherName,"zebrafish"))) { safef( chromStr, sizeof chromStr, "%sChrom" , otherName ); longXenoPsl1zoo2(tdb, item, otherName, chromStr ); } } +#ifdef OLD struct chain *getChainFromRange(char *chainTable, char *chrom, int chromStart, int chromEnd) /* get a list of chains for a range */ { char chainTable_chrom[256]; struct dyString *dy = dyStringNew(128); struct chain *chainList = NULL; struct sqlConnection *conn = hAllocConn(database); safef(chainTable_chrom, 256, "%s_%s",chrom, chainTable); if (hTableExists(database, chainTable_chrom) ) { /* lookup chain if not stored */ char **row; struct sqlResult *sr = NULL; @@ -14789,30 +14877,31 @@ qEnd = qSize - qStart; qStart = tmp; } chain = NULL; if (chainId != 0) { chain = chainLoadIdRange(database, chainTable, chrom, chromStart, chromEnd, chainId); if (chain != NULL) slAddHead(&chainList, chain); } } sqlFreeResult(&sr); } return chainList; } +#endif /* OLD */ void htcPseudoGene(char *htcCommand, char *item) /* Interface for selecting & displaying alignments from axtInfo * for an item from a genePred table. */ { struct genePred *gp = NULL; struct axtInfo *aiList = NULL; struct axt *axtList = NULL; struct sqlResult *sr; char **row; char trackTemp[256]; char *track = cartString(cart, "o"); char *chrom = cartString(cart, "c"); char *name = cartOptionalString(cart, "i"); char *db2 = cartString(cart, "db2");