struct vcfFile *vcfTabixFileMayOpen(char *fileOrUrl, char *chrom, int start, int end,
int maxErr, int maxRecords)
/* Parse header and rows within the given position range from a VCF file that has been
* compressed and indexed by tabix into a vcfFile object; return NULL if or if file has
* no items in range.
* If maxErr not zero, then continue to parse until this number of error have been reached.
* A maxErr less than zero does not stop and reports all errors. */
{
struct lineFile *lf = lineFileTabixMayOpen(fileOrUrl, TRUE);
struct vcfFile *vcff = vcfFileHeaderFromLineFile(lf, maxErr);
if (vcff == NULL)
return NULL;
if (isNotEmpty(chrom) && start != end)
{
if (lineFileSetTabixRegion(lf, chrom, start, end))
vcfParseData(vcff, maxRecords);
}
return vcff;
}
struct vcfFile *vcfTabixFileMayOpen(char *fileOrUrl, char *chrom, int start, int end,
int maxErr, int maxRecords)
/* Open a VCF file that has been compressed and indexed by tabix and
* parse VCF header, or return NULL if unable. If chrom is non-NULL,
* seek to the position range and parse all lines in range into
* vcff->records. If maxErr >= zero, then continue to parse until
* there are maxErr+1 errors. A maxErr less than zero does not stop
* and reports all errors. Set maxErr to VCF_IGNORE_ERRS for silence */
{
struct lineFile *lf = lineFileTabixMayOpen(fileOrUrl, TRUE);
struct vcfFile *vcff = vcfFileHeaderFromLineFile(lf, maxErr);
if (vcff == NULL)
return NULL;
if (isNotEmpty(chrom) && start != end)
{
if (lineFileSetTabixRegion(lf, chrom, start, end))
{
vcff->records = vcfParseData(vcff, maxRecords);
lineFileClose(&(vcff->lf)); // Not sure why it is closed. Angie?
}
}
return vcff;
}
struct slName *randomVcfIds(char *table, struct sqlConnection *conn, int count, boolean isTabix)
/* Return some semi-random IDs from a VCF file. */
{
/* Read 10000 items from vcf file, or if they ask for a big list, then 4x what they ask for. */
struct trackDb *tdb = hashFindVal(fullTableToTdbHash, table);
char *fileName = vcfFileName(tdb, conn, table, hDefaultChrom(database));
struct lineFile *lf = isTabix ? lineFileTabixMayOpen(fileName, TRUE) :
lineFileMayOpen(fileName, TRUE);
if (lf == NULL)
noWarnAbort();
int orderedCount = count * 4;
if (orderedCount < 100)
orderedCount = 100;
struct slName *idList = NULL;
char *words[4];
int i;
for (i = 0; i < orderedCount && lineFileChop(lf, words); i++)
{
// compress runs of identical ID, in case most are placeholder
if (i == 0 || !sameString(words[2], idList->name))
slAddHead(&idList, slNameNew(words[2]));
}
lineFileClose(&lf);
/* Shuffle list and trim it to count if necessary. */
shuffleList(&idList);
struct slName *sl;
for (sl = idList, i = 0; sl != NULL; sl = sl->next, i++)
{
if (i+1 >= count)
{
slNameFreeList(&(sl->next));
break;
}
}
freez(&fileName);
return idList;
}
void showSchemaVcf(char *table, struct trackDb *tdb, boolean isTabix)
/* Show schema on vcf. */
{
struct sqlConnection *conn = hAllocConn(database);
char *fileName = vcfFileName(tdb, conn, table, hDefaultChrom(database));
struct asObject *as = vcfAsObj();
hPrintf("<B>Database:</B> %s", database);
hPrintf(" <B>Primary Table:</B> %s<br>", table);
hPrintf("<B>VCF File:</B> %s", fileName);
hPrintf("<BR>\n");
hPrintf("<B>Format description:</B> %s<BR>", as->comment);
hPrintf("See the <A HREF=\"%s\" target=_blank>Variant Call Format specification</A> for more details<BR>\n",
"http://www.1000genomes.org/wiki/analysis/vcf4.0");
/* Put up table that describes fields. */
hTableStart();
hPrintf("<TR><TH>field</TH>");
hPrintf("<TH>description</TH> ");
puts("</TR>\n");
struct asColumn *col;
int colCount = 0;
for (col = as->columnList; col != NULL; col = col->next)
{
hPrintf("<TR><TD><TT>%s</TT></TD>", col->name);
hPrintf("<TD>%s</TD></TR>", col->comment);
++colCount;
}
hTableEnd();
/* Put up another section with sample rows. */
webNewSection("Sample Rows");
hTableStart();
/* Fetch sample rows. */
struct lineFile *lf = isTabix ? lineFileTabixMayOpen(fileName, TRUE) :
lineFileMayOpen(fileName, TRUE);
if (lf == NULL)
noWarnAbort();
char *row[VCF_MAX_SCHEMA_COLS];
int i;
for (i = 0; i < 10; i++)
{
int colCount = lineFileChop(lf, row);
int colIx;
if (i == 0)
{
// Print field names as column headers, using colCount to compute genotype span
hPrintf("<TR>");
for (colIx = 0, col = as->columnList; col != NULL && colIx < colCount;
colIx++, col = col->next)
{
if (sameString("genotypes", col->name) && colCount > colIx+1)
hPrintf("<TH colspan=%d>%s</TH>", colCount - colIx, col->name);
else
hPrintf("<TH>%s</TH>", col->name);
}
hPrintf("</TR>\n");
}
hPrintf("<TR>");
for (colIx=0; colIx < colCount; ++colIx)
{
if (colCount > VCFDATALINE_NUM_COLS && colIx == colCount - 1)
hPrintf("<TD>...</TD>");
else
writeHtmlCell(row[colIx]);
}
hPrintf("</TR>\n");
}
hTableEnd();
printTrackHtml(tdb);
/* Clean up and go home. */
lineFileClose(&lf);
freeMem(fileName);
hFreeConn(&conn);
}
