static boolean getMinQual(struct trackDb *tdb, double *retMinQual)
/* Return TRUE and set retMinQual if cart contains minimum QUAL filter */
{
if (cartOrTdbBoolean(cart, tdb, VCF_APPLY_MIN_QUAL_VAR, VCF_DEFAULT_APPLY_MIN_QUAL))
{
if (retMinQual != NULL)
*retMinQual = cartOrTdbDouble(cart, tdb, VCF_MIN_QUAL_VAR, VCF_DEFAULT_MIN_QUAL);
return TRUE;
}
return FALSE;
}
static void vcfTabixLoadItems(struct track *tg)
/* Load items in window from VCF file using its tabix index file. */
{
char *fileOrUrl = NULL;
/* Figure out url or file name. */
if (tg->parallelLoading)
{
/* do not use mysql during parallel-fetch load */
fileOrUrl = trackDbSetting(tg->tdb, "bigDataUrl");
}
else
{
struct sqlConnection *conn = hAllocConnTrack(database, tg->tdb);
fileOrUrl = bbiNameFromSettingOrTableChrom(tg->tdb, conn, tg->table, chromName);
hFreeConn(&conn);
}
if (isEmpty(fileOrUrl))
return;
int vcfMaxErr = -1;
struct vcfFile *vcff = NULL;
boolean hapClustEnabled = cartOrTdbBoolean(cart, tg->tdb, VCF_HAP_ENABLED_VAR, TRUE);
/* protect against temporary network error */
struct errCatch *errCatch = errCatchNew();
if (errCatchStart(errCatch))
{
vcff = vcfTabixFileMayOpen(fileOrUrl, chromName, winStart, winEnd, vcfMaxErr, -1);
if (vcff != NULL)
{
filterRecords(vcff, tg->tdb);
if (hapClustEnabled && vcff->genotypeCount > 1 && vcff->genotypeCount < 3000 &&
(tg->visibility == tvPack || tg->visibility == tvSquish))
vcfHapClusterOverloadMethods(tg, vcff);
else
{
tg->items = vcfFileToPgSnp(vcff, tg->tdb);
// pgSnp bases coloring/display decision on count of items:
tg->customInt = slCount(tg->items);
}
// Don't vcfFileFree here -- we are using its string pointers!
}
}
errCatchEnd(errCatch);
if (errCatch->gotError || vcff == NULL)
{
if (isNotEmpty(errCatch->message->string))
tg->networkErrMsg = cloneString(errCatch->message->string);
tg->drawItems = bigDrawWarning;
tg->totalHeight = bigWarnTotalHeight;
}
errCatchFree(&errCatch);
}
static void vcfLoadItems(struct track *tg)
/* Load items in window from VCF file. */
{
int vcfMaxErr = -1;
struct vcfFile *vcff = NULL;
boolean hapClustEnabled = cartOrTdbBoolean(cart, tg->tdb, VCF_HAP_ENABLED_VAR, TRUE);
char *table = tg->table;
struct customTrack *ct = tg->customPt;
struct sqlConnection *conn;
if (ct == NULL)
conn = hAllocConnTrack(database, tg->tdb);
else
{
conn = hAllocConn(CUSTOM_TRASH);
table = ct->dbTableName;
}
char *vcfFile = bbiNameFromSettingOrTable(tg->tdb, conn, table);
hFreeConn(&conn);
/* protect against parse error */
struct errCatch *errCatch = errCatchNew();
if (errCatchStart(errCatch))
{
vcff = vcfFileMayOpen(vcfFile, chromName, winStart, winEnd, vcfMaxErr, -1, TRUE);
if (vcff != NULL)
{
filterRecords(vcff, tg->tdb);
if (hapClustEnabled && vcff->genotypeCount > 1 && vcff->genotypeCount < 3000 &&
(tg->visibility == tvPack || tg->visibility == tvSquish))
vcfHapClusterOverloadMethods(tg, vcff);
else
{
tg->items = vcfFileToPgSnp(vcff, tg->tdb);
// pgSnp bases coloring/display decision on count of items:
tg->customInt = slCount(tg->items);
}
// Don't vcfFileFree here -- we are using its string pointers!
}
}
errCatchEnd(errCatch);
if (errCatch->gotError || vcff == NULL)
{
if (isNotEmpty(errCatch->message->string))
tg->networkErrMsg = cloneString(errCatch->message->string);
tg->drawItems = bigDrawWarning;
tg->totalHeight = bigWarnTotalHeight;
}
errCatchFree(&errCatch);
}
static void vcfRecordDetails(struct trackDb *tdb, struct vcfRecord *rec)
/* Display the contents of a single line of VCF, assumed to be from seqName
* (using seqName instead of rec->chrom because rec->chrom might lack "chr"). */
{
printf("<B>Name:</B> %s<BR>\n", rec->name);
// Since these are variants, if it looks like a dbSNP or dbVar ID, provide a link:
if (regexMatch(rec->name, "^rs[0-9]+$"))
{
printf("<B>dbSNP:</B> ");
printDbSnpRsUrl(rec->name, "%s", rec->name);
puts("<BR>");
}
else if (regexMatch(rec->name, "^[en]ss?v[0-9]+$"))
{
printf("<B>dbVar:</B> ");
printf("<A HREF=\"http://www.ncbi.nlm.nih.gov/dbvar/variants/%s/\" "
"TARGET=_BLANK>%s</A><BR>\n", rec->name, rec->name);
}
printCustomUrl(tdb, rec->name, TRUE);
boolean hapClustEnabled = cartOrTdbBoolean(cart, tdb, VCF_HAP_ENABLED_VAR, TRUE);
if (hapClustEnabled)
{
static char *formName = "vcfCfgHapCenter";
printf("<FORM NAME=\"%s\" ACTION=\"%s\">\n", formName, hgTracksName());
cartSaveSession(cart);
vcfCfgHaplotypeCenter(cart, tdb, tdb->track, FALSE, rec->file, rec->name,
seqName, rec->chromStart, formName);
printf("</FORM>\n");
}
char leftBase = rec->alleles[0][0];
unsigned int vcfStart = vcfRecordTrimIndelLeftBase(rec);
boolean showLeftBase = (rec->chromStart == vcfStart+1);
(void)vcfRecordTrimAllelesRight(rec);
char *displayAls[rec->alleleCount];
makeDisplayAlleles(rec, showLeftBase, leftBase, 20, TRUE, FALSE, displayAls);
printPosOnChrom(seqName, rec->chromStart, rec->chromEnd, NULL, FALSE, rec->name);
printf("<B>Reference allele:</B> %s<BR>\n", displayAls[0]);
vcfAltAlleleDetails(rec, displayAls);
vcfQualDetails(rec);
vcfFilterDetails(rec);
vcfInfoDetails(rec);
pgSnpCodingDetail(rec);
makeDisplayAlleles(rec, showLeftBase, leftBase, 5, FALSE, TRUE, displayAls);
vcfGenotypesDetails(rec, tdb, displayAls);
}
static void vcfGenotypesDetails(struct vcfRecord *rec, struct trackDb *tdb, char **displayAls)
/* Print summary of allele and genotype frequency, plus collapsible section
* with table of genotype details. */
{
struct vcfFile *vcff = rec->file;
if (vcff->genotypeCount == 0)
return;
// Wrapper table for collapsible section:
puts("<TABLE>");
pushWarnHandler(ignoreEm);
vcfParseGenotypes(rec);
popWarnHandler();
// Tally genotypes and alleles for summary:
int refs = 0, alts = 0, unks = 0;
int refRefs = 0, refAlts = 0, altAlts = 0, gtUnk = 0, gtOther = 0, phasedGts = 0;
int i;
for (i = 0; i < vcff->genotypeCount; i++)
{
struct vcfGenotype *gt = &(rec->genotypes[i]);
if (gt->isPhased)
phasedGts++;
if (gt->hapIxA == 0)
refs++;
else if (gt->hapIxA > 0)
alts++;
else
unks++;
if (!gt->isHaploid)
{
if (gt->hapIxB == 0)
refs++;
else if (gt->hapIxB > 0)
alts++;
else
unks++;
if (gt->hapIxA == 0 && gt->hapIxB == 0)
refRefs++;
else if (gt->hapIxA == 1 && gt->hapIxB == 1)
altAlts++;
else if ((gt->hapIxA == 1 && gt->hapIxB == 0) ||
(gt->hapIxA == 0 && gt->hapIxB == 1))
refAlts++;
else if (gt->hapIxA < 0 || gt->hapIxB < 0)
gtUnk++;
else
gtOther++;
}
}
printf("<B>Genotype count:</B> %d", vcff->genotypeCount);
if (differentString(seqName, "chrY"))
printf(" (%d phased)", phasedGts);
else
printf(" (haploid)");
puts("<BR>");
int totalAlleles = refs + alts + unks;
double refAf = (double)refs/totalAlleles;
double altAf = (double)alts/totalAlleles;
printf("<B>Alleles:</B> %s: %d (%.3f%%); %s: %d (%.3f%%)",
displayAls[0], refs, 100*refAf, displayAls[1], alts, 100*altAf);
if (unks > 0)
printf("; unknown: %d (%.3f%%)", unks, 100 * (double)unks/totalAlleles);
puts("<BR>");
// Should be a better way to detect haploid chromosomes than comparison with "chrY":
if (vcff->genotypeCount > 1 && differentString(seqName, "chrY"))
{
printf("<B>Genotypes:</B> %s/%s: %d (%.3f%%); %s/%s: %d (%.3f%%); %s/%s: %d (%.3f%%)",
displayAls[0], displayAls[0], refRefs, 100*(double)refRefs/vcff->genotypeCount,
displayAls[0], displayAls[1], refAlts, 100*(double)refAlts/vcff->genotypeCount,
displayAls[1], displayAls[1], altAlts, 100*(double)altAlts/vcff->genotypeCount);
if (gtUnk > 0)
printf("; unknown: %d (%.3f%%)", gtUnk, 100*(double)gtUnk/vcff->genotypeCount);
if (gtOther > 0)
printf("; other: %d (%.3f%%)", gtOther, 100*(double)gtOther/vcff->genotypeCount);
printf("<BR>\n");
if (rec->alleleCount == 2)
{
boolean showHW = cartOrTdbBoolean(cart, tdb, VCF_SHOW_HW_VAR, FALSE);
if (showHW)
printf("<B><A HREF=\"http://en.wikipedia.org/wiki/Hardy%%E2%%80%%93Weinberg_principle\" "
"TARGET=_BLANK>Hardy-Weinberg equilibrium</A>:</B> "
"P(%s/%s) = %.3f%%; P(%s/%s) = %.3f%%; P(%s/%s) = %.3f%%<BR>",
displayAls[0], displayAls[0], 100*refAf*refAf,
displayAls[0], displayAls[1], 100*2*refAf*altAf,
displayAls[1], displayAls[1], 100*altAf*altAf);
}
}
puts("<BR>");
vcfGenotypeTable(rec, tdb->track, displayAls);
puts("</TABLE>");
}
