static void wrapHgGeneLink(struct sqlConnection *conn, char *name,
char *label, char *geneTable)
/* Wrap label with link to hgGene if possible. */
{
char query[256];
struct sqlResult *sr;
char **row;
int rowOffset = hOffsetPastBin(database, seqName, "sgdGene");
sqlSafef(query, sizeof(query),
"select * from %s where name = '%s'", geneTable, name);
sr = sqlGetResult(conn, query);
if ((row = sqlNextRow(sr)) != NULL)
{
struct genePred *gp = genePredLoad(row+rowOffset);
printf("<A HREF=\"../cgi-bin/hgGene?db=%s", database);
printf("&hgg_gene=%s", gp->name);
printf("&hgg_chrom=%s", gp->chrom);
printf("&hgg_start=%d", gp->txStart);
printf("&hgg_end=%d", gp->txEnd);
printf("\">");
printf("%s", label);
printf("</A>");
}
else
printf("%s", label);
sqlFreeResult(&sr);
}
void doTriangle(struct trackDb *tdb, char *item, char *motifTable)
/* Display detailed info on a regulatory triangle item. */
{
int start = cartInt(cart, "o");
struct dnaSeq *seq = NULL;
struct dnaMotif *motif = loadDnaMotif(item, motifTable);
char *table = tdb->table;
int rowOffset = hOffsetPastBin(database, seqName, table);
char query[256];
struct sqlResult *sr;
char **row;
struct bed *hit = NULL;
struct sqlConnection *conn = hAllocConn(database);
cartWebStart(cart, database, "Regulatory Motif Info");
genericBedClick(conn, tdb, item, start, 6);
sqlSafef(query, sizeof query,
"select * from %s where name = '%s' and chrom = '%s' and chromStart = %d",
table, item, seqName, start);
sr = sqlGetResult(conn, query);
row = sqlNextRow(sr);
if (row != NULL)
hit = bedLoadN(row + rowOffset, 6);
sqlFreeResult(&sr);
if (hit != NULL)
{
seq = hDnaFromSeq(database, hit->chrom, hit->chromStart, hit->chromEnd, dnaLower);
if (hit->strand[0] == '-')
reverseComplement(seq->dna, seq->size);
}
motifHitSection(seq, motif);
printTrackHtml(tdb);
}
void doPeakClusters(struct trackDb *tdb, char *item)
/* Display detailed info about a cluster of DNase peaks from other tracks. */
{
int start = cartInt(cart, "o");
char *table = tdb->table;
int rowOffset = hOffsetPastBin(database, seqName, table);
char query[256];
struct sqlResult *sr;
char **row;
struct bed *cluster = NULL;
struct sqlConnection *conn = hAllocConn(database);
cartWebStart(cart, database, "%s item details", tdb->shortLabel);
sqlSafef(query, sizeof(query),
"select * from %s where name = '%s' and chrom = '%s' and chromStart = %d",
table, item, seqName, start);
sr = sqlGetResult(conn, query);
row = sqlNextRow(sr);
if (row != NULL)
cluster = bedLoadN(row+rowOffset, 5);
sqlFreeResult(&sr);
if (cluster != NULL)
{
/* Get list of subgroups to display */
char *inputTableFieldDisplay = trackDbSetting(tdb, "inputTableFieldDisplay");
if (inputTableFieldDisplay != NULL)
{
struct slName *fieldList = stringToSlNames(inputTableFieldDisplay);
char *inputTrackTable = trackDbRequiredSetting(tdb, "inputTrackTable");
/* Print out some information about the cluster overall. */
printf("<B>Items in Cluster:</B> %s of %d<BR>\n", cluster->name,
sqlRowCount(conn, sqlCheckIdentifier(inputTrackTable)));
printf("<B>Cluster Score (out of 1000):</B> %d<BR>\n", cluster->score);
printPos(cluster->chrom, cluster->chromStart, cluster->chromEnd, NULL, TRUE, NULL);
/* In a new section put up list of hits. */
webNewSection("List of Items in Cluster");
webPrintLinkTableStart();
printClusterTableHeader(fieldList, FALSE, FALSE, TRUE);
printPeakClusterInfo(tdb, cart, conn, inputTrackTable, fieldList, cluster);
}
else
errAbort("Missing required trackDb setting %s for track %s",
"inputTableFieldDisplay", tdb->track);
webPrintLinkTableEnd();
}
printf("<A HREF=\"%s&g=htcListItemsAssayed&table=%s\" TARGET_blank>", hgcPathAndSettings(),
tdb->track);
printf("List all items assayed");
printf("</A><BR>\n");
webNewSection("Track Description");
printTrackHtml(tdb);
hFreeConn(&conn);
}
void doTransRegCode(struct trackDb *tdb, char *item, char *motifTable)
/* Display detailed info on a transcriptional regulatory code item. */
{
struct dnaMotif *motif = loadDnaMotif(item, motifTable);
int start = cartInt(cart, "o");
struct dnaSeq *seq = NULL;
char *table = tdb->table;
int rowOffset = hOffsetPastBin(database, seqName, table);
char query[256];
struct sqlResult *sr;
char **row;
struct sqlConnection *conn = hAllocConn(database);
struct transRegCode *trc = NULL;
cartWebStart(cart, database, "Regulatory Code Info");
sqlSafef(query, sizeof query,
"select * from %s where name = '%s' and chrom = '%s' and chromStart = %d",
table, item, seqName, start);
sr = sqlGetResult(conn, query);
row = sqlNextRow(sr);
if (row != NULL)
trc = transRegCodeLoad(row+rowOffset);
sqlFreeResult(&sr);
if (trc != NULL)
{
char strand[2];
seq = hDnaFromSeq(database, trc->chrom, trc->chromStart, trc->chromEnd, dnaLower);
if (seq->size != motif->columnCount)
{
printf("WARNING: seq->size = %d, motif->colCount=%d<BR>\n",
seq->size, motif->columnCount);
strand[0] = '?';
seq = NULL;
}
else
{
strand[0] = dnaMotifBestStrand(motif, seq->dna);
if (strand[0] == '-')
reverseComplement(seq->dna, seq->size);
}
strand[1] = 0;
printf("<B>Name:</B> ");
sacCerHgGeneLinkName(conn, trc->name);
printf("<BR>\n");
printf("<B>ChIP-chip Evidence:</B> %s<BR>\n", trc->chipEvidence);
printf("<B>Species conserved in:</B> %d of 2<BR>\n", trc->consSpecies);
if (seq != NULL)
printf("<B>Bit Score of Motif Hit:</B> %4.2f<BR>\n",
dnaMotifBitScore(motif, seq->dna));
printf("<B>Item score:</B> %d<BR>\n", trc->score);
printPosOnChrom(trc->chrom, trc->chromStart, trc->chromEnd, strand, TRUE, trc->name);
}
motifHitSection(seq, motif);
printTrackHtml(tdb);
}
static void getAligns(struct sqlConnection *conn, struct hash *refSeqVerInfoTbl, char *outFile)
/* get request alignments from database */
{
int off = hOffsetPastBin(sqlGetDatabase(conn), NULL, "refSeqAli");
struct sqlResult *sr = sqlGetResult(conn, "SELECT * FROM refSeqAli");
FILE *fh = mustOpen(outFile, "w");
char **row;
while ((row = sqlNextRow(sr)) != NULL)
{
struct psl *psl = pslLoad(row+off);
processPsl(fh, refSeqVerInfoTbl, psl);
pslFree(&psl);
}
carefulClose(&fh);
sqlFreeResult(&sr);
}
void getSeqGapsUnsplit(struct sqlConnection *conn, struct hash *chromHash)
/* Return a tree of ranges for sequence gaps in all chromosomes,
* assuming an unsplit gap table -- when the table is unsplit, it's
* probably for a scaffold assembly where we *really* don't want
* to do one query per scaffold! */
{
struct rbTreeNode **stack = lmAlloc(qLm, 256 * sizeof(stack[0]));
struct rbTree *tree = rbTreeNewDetailed(simpleRangeCmp, qLm, stack);
int rowOffset = hOffsetPastBin(sqlGetDatabase(conn), NULL, "gap");
struct sqlResult *sr;
char **row;
char *prevChrom = NULL;
sr = sqlGetResult(conn, "NOSQLINJ select * from gap order by chrom");
while ((row = sqlNextRow(sr)) != NULL)
{
struct agpGap gap;
struct simpleRange *range;
agpGapStaticLoad(row+rowOffset, &gap);
if (prevChrom == NULL)
prevChrom = cloneString(gap.chrom);
else if (! sameString(prevChrom, gap.chrom))
{
setNGap(prevChrom, chromHash, tree);
freeMem(prevChrom);
stack = lmAlloc(qLm, 256 * sizeof(stack[0]));
tree = rbTreeNewDetailed(simpleRangeCmp, qLm, stack);
prevChrom = cloneString(gap.chrom);
}
lmAllocVar(tree->lm, range);
range->start = gap.chromStart;
range->end = gap.chromEnd;
rbTreeAdd(tree, range);
}
if (prevChrom != NULL)
{
setNGap(prevChrom, chromHash, tree);
freeMem(prevChrom);
}
sqlFreeResult(&sr);
}
void intronSizes(char *database, char *table)
/* intronSizes - Output list of intron sizes.. */
{
struct dyString *query = newDyString(1024);
struct sqlConnection *conn;
struct sqlResult *sr;
char **row;
struct genePred *gp;
int rowOffset;
struct bed *bedList = NULL, *bed = NULL;
hSetDb(database);
rowOffset = hOffsetPastBin(NULL, table);
conn = hAllocConn(database);
sqlDyStringPrintf(query, "select * from %s", table);
if (chromName != NULL)
dyStringPrintf(query, " where chrom = '%s'", chromName);
if (cgiBoolean("withUtr"))
{
dyStringPrintf(query, " %s txStart != cdsStart",
(chromName == NULL ? "where" : "and"));
}
sr = sqlGetResult(conn, query->string);
while ((row = sqlNextRow(sr)) != NULL)
{
gp = genePredLoad(row+rowOffset);
genePredIntrons(gp, &bedList);
slReverse(&bedList);
for (bed = bedList ; bed != NULL ; bed=bed->next)
bedTabOutN(bed,6, stdout);
bedFreeList(&bedList);
genePredFree(&gp);
}
sqlFreeResult(&sr);
hFreeConn(&conn);
}
void intronEnds(char *database, char *table)
/* intronEnds - Gather stats on intron ends.. */
{
struct dyString *query = newDyString(1024);
struct sqlConnection *conn;
struct sqlResult *sr;
char **row;
struct genePred *gp;
int total = 0;
int gtag = 0;
int gcag = 0;
int atac = 0;
int ctac = 0;
DNA ends[4];
int exonIx, txStart;
struct dnaSeq *seq;
int rowOffset;
char strand;
rowOffset = hOffsetPastBin(database, NULL, table);
conn = hAllocConn(database);
sqlDyStringPrintf(query, "select * from %s", table);
if (chromName != NULL)
dyStringPrintf(query, " where chrom = '%s'", chromName);
if (cgiBoolean("withUtr"))
{
dyStringPrintf(query, " %s txStart != cdsStart",
(chromName == NULL ? "where" : "and"));
}
sr = sqlGetResult(conn, query->string);
while ((row = sqlNextRow(sr)) != NULL)
{
gp = genePredLoad(row+rowOffset);
strand = gp->strand[0];
txStart = gp->txStart;
seq = hDnaFromSeq(database, gp->chrom, txStart, gp->txEnd, dnaLower);
for (exonIx=1; exonIx < gp->exonCount; ++exonIx)
{
++total;
memcpy(ends, seq->dna + gp->exonEnds[exonIx-1] - txStart, 2);
memcpy(ends+2, seq->dna + gp->exonStarts[exonIx] - txStart - 2, 2);
if (strand == '-')
reverseComplement(ends, 4);
if (ends[0] == 'g' && ends[1] == 't' && ends[2] == 'a' && ends[3] == 'g')
++gtag;
if (ends[0] == 'g' && ends[1] == 'c' && ends[2] == 'a' && ends[3] == 'g')
++gcag;
if (ends[0] == 'a' && ends[1] == 't' && ends[2] == 'a' && ends[3] == 'c')
++atac;
if (ends[0] == 'c' && ends[1] == 't' && ends[2] == 'a' && ends[3] == 'c')
++ctac;
}
freeDnaSeq(&seq);
genePredFree(&gp);
}
sqlFreeResult(&sr);
hFreeConn(&conn);
printf("gt/ag %d (%4.2f)\n", gtag, 100.0*gtag/total);
printf("gc/ag %d (%4.2f)\n", gcag, 100.0*gcag/total);
printf("at/ac %d (%4.2f)\n", atac, 100.0*atac/total);
printf("ct/ac %d (%4.2f)\n", ctac, 100.0*ctac/total);
printf("Total %d\n", total);
}
void loadSimpleBed(struct track *tg)
/* Load the items in one track - just move beds in
* window... */
{
struct bed *(*loader)(char **row);
struct bed *bed, *list = NULL;
char **row;
int rowOffset;
char *words[3];
int wordCt;
char query[128];
char *setting = NULL;
bool doScoreCtFilter = FALSE;
int scoreFilterCt = 0;
char *topTable = NULL;
if (tg->bedSize <= 3)
loader = bedLoad3;
else if (tg->bedSize == 4)
loader = bedLoad;
else if (tg->bedSize == 5)
loader = bedLoad5;
else
loader = bedLoad6;
// pairedTagAlign loader is required for base coloring using sequence from seq1 & seq2
// after removing optional bin column, this loader assumes seq1 and seq2 are in
// row[6] and row[7] respectively of the sql result.
if ((setting = trackDbSetting(tg->tdb, BASE_COLOR_USE_SEQUENCE))
&& sameString(setting, "seq1Seq2"))
loader = bedLoadPairedTagAlign;
/* limit to a specified count of top scoring items.
* If this is selected, it overrides selecting item by specified score */
if ((setting = trackDbSettingClosestToHome(tg->tdb, "filterTopScorers")) != NULL)
{
wordCt = chopLine(cloneString(setting), words);
if (wordCt >= 3)
{
doScoreCtFilter = cartUsualBooleanClosestToHome(cart, tg->tdb, FALSE,
"filterTopScorersOn",sameString(words[0], "on"));
scoreFilterCt = cartUsualIntClosestToHome(cart, tg->tdb, FALSE,
"filterTopScorersCt", atoi(words[1]));
topTable = words[2];
/* if there are not too many rows in the table then can define */
/* top table as the track or subtrack table */
if (sameWord(topTable, "self"))
topTable = cloneString(tg->table);
}
}
/* Get list of items */
if (tg->isBigBed)
{ // avoid opening an unneeded db connection for bigBed; required not to use mysql for parallel fetch tracks
char *scoreFilter = cartOrTdbString(cart, tg->tdb, "scoreFilter", NULL);
struct lm *lm = lmInit(0);
struct bigBedInterval *bb, *bbList = bigBedSelectRange(tg, chromName, winStart, winEnd, lm);
char *bedRow[32];
char startBuf[16], endBuf[16];
int minScore = 0;
if (scoreFilter)
minScore = atoi(scoreFilter);
for (bb = bbList; bb != NULL; bb = bb->next)
{
bigBedIntervalToRow(bb, chromName, startBuf, endBuf, bedRow, ArraySize(bedRow));
bed = loader(bedRow);
if (scoreFilter == NULL || bed->score >= minScore)
slAddHead(&list, bed);
}
lmCleanup(&lm);
}
else
{
struct sqlConnection *conn = hAllocConnTrack(database, tg->tdb);
struct sqlResult *sr = NULL;
/* limit to items above a specified score */
char *scoreFilterClause = getScoreFilterClause(cart, tg->tdb,NULL);
if (doScoreCtFilter && (topTable != NULL) && hTableExists(database, topTable))
{
sqlSafef(query, sizeof(query),"select * from %s order by score desc limit %d",
topTable, scoreFilterCt);
sr = sqlGetResult(conn, query);
rowOffset = hOffsetPastBin(database, hDefaultChrom(database), topTable);
}
else if(scoreFilterClause != NULL && tg->bedSize >= 5)
{
sr = hRangeQuery(conn, tg->table, chromName, winStart, winEnd, scoreFilterClause, &rowOffset);
}
else
{
sr = hRangeQuery(conn, tg->table, chromName, winStart, winEnd, NULL, &rowOffset);
}
freeMem(scoreFilterClause);
while ((row = sqlNextRow(sr)) != NULL)
{
bed = loader(row+rowOffset);
slAddHead(&list, bed);
}
sqlFreeResult(&sr);
hFreeConn(&conn);
}
if (doScoreCtFilter)
{
/* filter out items not in this window */
struct bed *newList = bedFilterListInRange(list, NULL, chromName, winStart, winEnd);
list = newList;
}
slReverse(&list);
tg->items = list;
}
void doGvf(struct trackDb *tdb, char *item)
/* Show details for variants represented as GVF, stored in a bed8Attrs table */
{
struct sqlConnection *conn = hAllocConn(database);
int start = cartInt(cart, "o");
char query[1024];
sqlSafef(query, sizeof(query),
"select * from %s where name = '%s' and chrom = '%s' and chromStart = %d",
tdb->table, item, seqName, start);
struct sqlResult *sr = sqlGetResult(conn, query);
char **row;
if ((row = sqlNextRow(sr)) == NULL)
errAbort("doGvfDetails: can't find item '%s' in %s at %s:%d", item, database, seqName, start);
int rowOffset = hOffsetPastBin(database, seqName, tdb->table);
struct bed8Attrs *ba = bed8AttrsLoad(row+rowOffset);
bedPrintPos((struct bed *)ba, 3, tdb);
int i = 0;
// Note: this loop modifies ba->attrVals[i], assuming we won't use them again:
for (i = 0; i < ba->attrCount; i++)
{
// The ID is the bed8Attrs name and has already been displayed:
if (sameString(ba->attrTags[i], "ID"))
continue;
cgiDecode(ba->attrVals[i], ba->attrVals[i], strlen(ba->attrVals[i]));
char *tag = ba->attrTags[i];
// User-defined keywords used in dbVar's GVF:
if (sameString(tag, "var_type")) // This one isn't anymore, but I add it back (hg18.txt).
tag = "Variant type";
else if (sameString(tag, "clinical_int"))
tag = "Clinical interpretation";
else if (sameString(tag, "var_origin"))
tag = "Variant origin";
else if (islower(tag[0]))
// Uppercase for nice display, assuming user doesn't care which keywords are
// user-defined vs. GVF standard:
tag[0] = toupper(tag[0]);
// GVF standard Start_range and End_range tags (1-based coords):
if (sameString(tag, "Start_range") || sameString(tag, "End_range"))
{
char *copy = cloneString(ba->attrVals[i]);
char *words[3];
int wordCount = chopCommas(copy, words);
if (wordCount == 2 &&
(sameString(".", words[0]) || isInteger(words[0])) &&
(sameString(".", words[1]) || isInteger(words[1])))
{
boolean isStartRange = sameString(tag, "Start_range");
char *rangeStart = words[0], *rangeEnd = words[1];
if (sameString(".", rangeStart))
rangeStart = "unknown";
if (sameString(".", rangeEnd))
rangeEnd = "unknown";
if (isStartRange)
printf("<B>Start range</B>: outer start %s, inner start %s<BR>\n",
rangeStart, rangeEnd);
else
printf("<B>End range</B>: inner end %s, outer end %s<BR>\n",
rangeStart, rangeEnd);
}
else
// not formatted as expected, just print as-is:
printf("<B>%s</B>: %s<BR>\n", tag, htmlEncode(ba->attrVals[i]));
}
// Parent sounds like mom or dad (as in var_origin)... tweak it too:
else if (sameString(tag, "Parent"))
{
printf("<B>Variant region:</B> "
"<A HREF=\"http://www.ncbi.nlm.nih.gov/dbvar/variants/%s/\" "
"TARGET=_BLANK>%s</A><BR>\n", ba->attrVals[i], htmlEncode(ba->attrVals[i]));
}
else if (sameString(tag, "Name"))
{
char *url = trackDbSetting(tdb, "url");
// Show the Name only if it hasn't already appeared in the URL:
if (url == NULL || !stringIn("$$", url))
printf("<B>%s</B>: %s<BR>\n", tag, htmlEncode(ba->attrVals[i]));
}
else if (sameWord(tag, "Phenotype_id") && startsWith("HPO:HP:", ba->attrVals[i]))
{
subChar(tag, '_', ' ');
printf("<B>%s</B>: <A HREF=\"http://www.berkeleybop.org/obo/%s\" "
"TARGET=_BLANK>%s</A><BR>\n", tag, ba->attrVals[i]+strlen("HPO:"),
htmlEncode(ba->attrVals[i]));
}
else
{
subChar(tag, '_', ' ');
printf("<B>%s</B>: %s<BR>\n", tag, htmlEncode(ba->attrVals[i]));
}
}
sqlFreeResult(&sr);
hFreeConn(&conn);
/* printTrackHtml is done in genericClickHandlerPlus. */
}
void doTransRegCodeProbe(struct trackDb *tdb, char *item,
char *codeTable, char *motifTable,
char *tfToConditionTable, char *conditionTable)
/* Display detailed info on a ChIP-chip probe from transRegCode experiments. */
{
char query[256];
struct sqlResult *sr;
char **row;
int rowOffset = hOffsetPastBin(database, seqName, tdb->table);
struct sqlConnection *conn = hAllocConn(database);
struct transRegCodeProbe *probe = NULL;
cartWebStart(cart, database, "ChIP-chip Probe Info");
sqlSafef(query, sizeof(query), "select * from %s where name = '%s'",
tdb->table, item);
sr = sqlGetResult(conn, query);
if ((row = sqlNextRow(sr)) != NULL)
probe = transRegCodeProbeLoad(row+rowOffset);
sqlFreeResult(&sr);
if (probe != NULL)
{
struct tfData *tfList = NULL, *tf;
struct hash *tfHash = newHash(0);
struct transRegCode *trc;
int i;
/* Print basic info. */
printf("<B>Name:</B> %s<BR>\n", probe->name);
printPosOnChrom(probe->chrom, probe->chromStart, probe->chromEnd,
NULL, TRUE, probe->name);
/* Make up list of all transcriptionFactors. */
for (i=0; i<probe->tfCount; ++i)
{
/* Parse out factor and condition. */
char *tfName = probe->tfList[i];
char *condition = strchr(tfName, '_');
struct tfCond *cond;
if (condition != NULL)
*condition++ = 0;
else
condition = "n/a";
tf = hashFindVal(tfHash, tfName);
if (tf == NULL)
{
AllocVar(tf);
hashAddSaveName(tfHash, tfName, tf, &tf->name);
slAddHead(&tfList, tf);
}
AllocVar(cond);
cond->name = cloneString(condition);
cond->binding = probe->bindVals[i];
slAddHead(&tf->conditionList, cond);
}
slSort(&tfList, tfDataCmpName);
/* Fold in motif hits in region. */
if (sqlTableExists(conn, codeTable))
{
sr = hRangeQuery(conn, codeTable,
probe->chrom, probe->chromStart, probe->chromEnd,
"chipEvidence != 'none'", &rowOffset);
while ((row = sqlNextRow(sr)) != NULL)
{
trc = transRegCodeLoad(row+rowOffset);
tf = hashFindVal(tfHash, trc->name);
if (tf != NULL)
slAddTail(&tf->trcList, trc);
}
sqlFreeResult(&sr);
}
if (tfList == NULL)
printf("No significant immunoprecipitation.");
else
{
tfBindLevelSection(tfList, conn, motifTable, tfToConditionTable);
}
transRegCodeProbeFree(&probe);
growthConditionSection(conn, conditionTable);
}
printf("\n<HR>\n");
printTrackHtml(tdb);
hFreeConn(&conn);
}
