5f7b28612df1ff4e29c31020b33f8fee9c097b11
chmalee
Tue Oct 10 12:35:00 2023 -0700
Check for refseq historical old table existence before querying them for hgvs terms, refs #26016
diff --git src/hg/lib/hgHgvs.c src/hg/lib/hgHgvs.c
index 415becd..883da03 100644
--- src/hg/lib/hgHgvs.c
+++ src/hg/lib/hgHgvs.c
@@ -728,31 +728,31 @@
nmAcc);
}
else if (hTableExists(db, "refGene"))
{
// Trim .version if present since our genbank tables don't use versioned names.
char *trimmed = cloneFirstWordByDelimiter(nmAcc, '.');
sqlSafef(query, sizeof(query), "select l.protAcc from %s l, refGene r "
"where r.name = '%s' and l.mrnaAcc = r.name "
"and l.protAcc != '' order by length(l.protAcc), l.protAcc",
refLinkTable, trimmed);
}
else return NULL;
struct sqlConnection *conn = hAllocConn(db);
char *npAcc = sqlQuickString(conn, query);
// if user passed in old versioned transcript, check in *Old tables:
-if (npAcc == NULL && hDbHasNcbiRefSeq(db) && strchr(nmAcc, '.'))
+if (npAcc == NULL && hDbHasNcbiRefSeqOld(db) && strchr(nmAcc, '.'))
{
sqlSafef(query, sizeof(query), "select protAcc from ncbiRefSeqLinkOld where id = '%s'", nmAcc);
npAcc = sqlQuickString(conn, query);
}
hFreeConn(&conn);
return npAcc;
}
static char *lrgProteinToTx(char *db, char *protAcc)
/* Return the LRG_ transcript accession for protAcc. Each LRG_NpM has a corresponding LRG_NtM. */
{
int accLen = strlen(protAcc);
char txAcc[accLen+1];
safecpy(txAcc, sizeof(txAcc), protAcc);
char *p = strrchr(txAcc, 'p');
@@ -857,31 +857,31 @@
struct sqlConnection *conn = hAllocConn(db);
seq = sqlQuickString(conn, query);
hFreeConn(&conn);
freeMem(txAcc);
}
}
else
{
if (hDbHasNcbiRefSeq(db))
{
char query[2048];
sqlSafef(query, sizeof(query), "select seq from ncbiRefSeqPepTable "
"where name = '%s'", acc);
struct sqlConnection *conn = hAllocConn(db);
seq = sqlQuickString(conn, query);
- if (seq == NULL)
+ if (seq == NULL && hDbHasNcbiRefSeqOld(db))
{
sqlSafef(query, sizeof(query), "select seq from ncbiRefSeqPepTableOld "
"where name = '%s'", acc);
seq = sqlQuickString(conn, query);
}
hFreeConn(&conn);
}
else
{
aaSeq *aaSeq = hGenBankGetPep(db, acc, NULL);
if (aaSeq)
seq = aaSeq->dna;
}
}
return seq;
@@ -1218,31 +1218,31 @@
seq = sqlQuickString(conn, query);
hFreeConn(&conn);
}
}
else if (startsWith("ENS", acc ))
{
// Construct it from the genome I guess?
seq = getGencodeSeq(db, acc);
}
else
{
struct dnaSeq *cdnaSeq = NULL;
if (hDbHasNcbiRefSeq(db))
{
cdnaSeq = hDnaSeqGet(db, acc, "seqNcbiRefSeq", "extNcbiRefSeq");
- if (cdnaSeq == NULL)
+ if (cdnaSeq == NULL && hDbHasNcbiRefSeqOld(db))
cdnaSeq = hDnaSeqGet(db, acc, "seqNcbiRefSeqOld", "extNcbiRefSeqOld");
}
else
cdnaSeq = hGenBankGetMrna(db, acc, NULL);
if (cdnaSeq)
seq = dnaSeqCannibalize(&cdnaSeq);
}
return seq;
}
static boolean getCds(char *db, char *acc, struct genbankCds *retCds)
/* Get the CDS info for genbank/LRG/ENS acc; return FALSE if not found or not applicable. */
{
if (trackHubDatabase(db))
return FALSE;
@@ -1264,31 +1264,31 @@
if (startsWith("LRG_", acc))
sqlSafef(query, sizeof(query),
"select cds from lrgCds where id = '%s'", acc);
else if (hDbHasNcbiRefSeq(db) &&
// This is a hack to allow us to fall back on refSeqAli if ncbiRefSeqPsl is incomplete:
strchr(acc, '.'))
sqlSafef(query, sizeof(query),
"select cds from ncbiRefSeqCds where id = '%s'", acc);
else
sqlSafef(query, sizeof(query),
"SELECT c.name FROM %s as c, %s as g WHERE (g.acc = '%s') AND "
"(g.cds != 0) AND (g.cds = c.id)", cdsTable, gbCdnaInfoTable, acc);
struct sqlConnection *conn = hAllocConn(db);
char cdsBuf[2048];
cdsStr = sqlQuickQuery(conn, query, cdsBuf, sizeof(cdsBuf));
- if (isEmpty(cdsStr) && strchr(acc, '.'))
+ if (isEmpty(cdsStr) && strchr(acc, '.') && hDbHasNcbiRefSeqOld(db))
{
sqlSafef(query, sizeof(query),
"select cds from ncbiRefSeqCdsOld where id = '%s'", acc);
cdsStr = sqlQuickQuery(conn, query, cdsBuf, sizeof(cdsBuf));
}
hFreeConn(&conn);
}
if (isNotEmpty(cdsStr))
gotCds = (genbankCdsParse(cdsStr, retCds) &&
retCds->startComplete && retCds->start != retCds->end);
return gotCds;
}
static char refBaseFromProt(char *change)
/* If change starts with an amino acid 3-letter or 1-letter code then return the 1-letter code,
@@ -1905,31 +1905,31 @@
}
}
}
else
{
char *pslTable = pslTableForAcc(db, acc);
if (pslTable && hTableExists(db, pslTable))
{
if (hgvs->type == hgvstCoding)
getCds(db, acc, cds);
txAli = pslForQName(db, pslTable, acc);
}
// try the old alignments if we can't find it in the current alignments
if (!txAli && sameString(pslTable, "ncbiRefSeqPsl"))
{
- if (hTableExists(db, "ncbiRefSeqPslOld"))
+ if (hDbHasNcbiRefSeqOld(db))
{
txAli = pslForQName(db, "ncbiRefSeqPslOld", acc);
if (txAli)
pslTable = "ncbiRefSeqPslOld";
}
// As of 9/26/16, ncbiRefSeqPsl is missing some items (#13673#note-443) -- so fall back
// on UCSC alignments.
if (!txAli && hTableExists(db, "refSeqAli"))
{
char *accNoVersion = cloneFirstWordByDelimiter(acc, '.');
if (hgvs->type == hgvstCoding)
getCds(db, accNoVersion, cds);
txAli = pslForQName(db, "refSeqAli", accNoVersion);
if (txAli)
{
@@ -1994,31 +1994,31 @@
char *txAcc = NULL;
if (startsWith("LRG_", acc))
txAcc = lrgProteinToTx(db, acc);
else if (startsWith("ENS", acc))
txAcc = gencodeProteinToTx(db, acc);
else if (startsWith("NP_", acc) || startsWith("XP_", acc))
{
struct sqlConnection *conn = hAllocConn(db);
char query[2048];
if (hDbHasNcbiRefSeq(db))
{
sqlSafef(query, sizeof(query), "select mrnaAcc from ncbiRefSeqLink where protAcc = '%s'",
acc);
txAcc = sqlQuickString(conn, query);
// user may have passed previous versioned transcript, check the *Old tables:
- if (!txAcc)
+ if (!txAcc && hDbHasNcbiRefSeqOld(db))
{
sqlSafef(query, sizeof(query), "select mrnaAcc from ncbiRefSeqLinkOld where protAcc = '%s'",
acc);
txAcc = sqlQuickString(conn, query);
}
}
else if (hTableExists(db, "refGene"))
{
sqlSafef(query, sizeof(query), "select mrnaAcc from %s l, refGene r "
"where l.protAcc = '%s' and r.name = l.mrnaAcc", refLinkTable, acc);
txAcc = sqlQuickString(conn, query);
}
else
return NULL;
hFreeConn(&conn);