struct visiMatch *visiSearch(struct sqlConnection *conn, char *searchString)
/* visiSearch - return list of images that match searchString sorted
* by how well they match. This will search most fields in the
* database. */
{
struct visiMatch *matchList;
struct slName *wordList = stringToSlNames(searchString);
int wordCount = slCount(wordList);
struct visiSearcher *searcher = visiSearcherNew(wordCount);
visiGeneMatchContributor(searcher, conn, wordList);
visiGeneMatchYear(searcher, conn, wordList);
visiGeneMatchProbeId(searcher, conn, wordList);
visiGeneMatchGene(searcher, conn, wordList);
visiGeneMatchAccession(searcher, conn, wordList);
visiGeneMatchBodyPart(searcher, conn, wordList);
visiGeneMatchSex(searcher, conn, wordList);
visiGeneMatchStage(searcher, conn, wordList);
visiGeneMatchSubmitId(searcher, conn, wordList);
visiGeneMatchOrganism(searcher, conn, wordList);
visiGeneMatchDescription(searcher, conn, wordList);
matchList = visiSearcherSortResults(searcher, conn);
searcher->matchList = NULL; /* Transferring memory ownership to return val. */
visiSearcherFree(&searcher);
slFreeList(&wordList);
return matchList;
}
void mafOrtholog(char *database, char *track, char *genePredFile, char *outFile)
/* mafOrtholog - find orthlogs in other species based on maf alignment and reference genePred */
{
struct slName *orgList = NULL;
FILE *f = mustOpen(outFile, "w");
struct genePredReader *gpr = genePredReaderFile(genePredFile, NULL);
struct genePred *gpList = genePredReaderAll(gpr), *gp = NULL;
struct sqlConnection *conn = hAllocConn();
if (optionExists("nibDir"))
nibDir = optionVal("nibDir", NULL);
if (optionExists("orgs"))
{
char *orgFile = optionVal("orgs", NULL);
char *buf;
readInGulp(orgFile, &buf, NULL);
orgList = stringToSlNames(buf);
}
for (gp = gpList ; gp != NULL ; gp=gp->next)
{
struct mafAli *maf = NULL;
if (thickOnly)
maf = mafLoadInRegion(conn, track,
gp->chrom, gp->cdsStart, gp->cdsEnd);
else
maf = mafLoadInRegion(conn, track,
gp->chrom, gp->txStart, gp->txEnd);
if (meFirst)
moveMeToFirst(maf, database);
printOrthologs(f, maf, gp);
mafAliFree(&maf);
}
carefulClose(&f);
}
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);
}
static void mafFrags(char *database, char *track, char *bedFile, char *mafFile)
/* mafFrags - Collect MAFs from regions specified in a 6 column bed file. */
{
struct slName *orgList = NULL;
struct lineFile *lf = lineFileOpen(bedFile, TRUE);
FILE *f = mustOpen(mafFile, "w");
if (optionExists("orgs"))
{
char *orgFile = optionVal("orgs", NULL);
char *buf;
readInGulp(orgFile, &buf, NULL);
orgList = stringToSlNames(buf);
/* Ensure that org list starts with database. */
struct slName *me = slNameFind(orgList, database);
if (me == NULL)
errAbort("Need to have reference database '%s' in %s", database, orgFile);
if (me != orgList)
{
slRemoveEl(&orgList, me);
slAddHead(&orgList, me);
}
}
mafWriteStart(f, "zero");
if (bed12)
{
char *row[12];
while (lineFileRow(lf, row))
{
struct bed *bed = bedLoadN(row, ArraySize(row));
struct mafAli *maf = mafFromBed12(database, track, bed, orgList);
if (meFirst)
moveMeToFirst(maf, bed->name);
mafWrite(f, maf);
mafAliFree(&maf);
bedFree(&bed);
}
}
else
{
char *row[6];
while (lineFileRow(lf, row))
{
struct bed *bed = bedLoadN(row, ArraySize(row));
processBed6(database, track, f, bed, orgList);
bedFree(&bed);
}
}
mafWriteEnd(f);
carefulClose(&f);
}
struct slName *keyFileList(struct column *col)
/* Make up list from key file for this column.
* return NULL if no key file. */
{
char *fileName = keyFileName(col);
char *buf;
struct slName *list;
if (fileName == NULL)
return NULL;
readInGulp(fileName, &buf, NULL);
list = stringToSlNames(buf);
freez(&buf);
return list;
}
static struct genePos *associationAdvFilter(struct column *col,
struct sqlConnection *conn, struct genePos *list)
/* Do advanced filter on position. */
{
char *terms = advFilterVal(col, "terms");
if (terms != NULL)
{
boolean orLogic = advFilterOrLogic(col, "logic", TRUE);
struct slName *termList = stringToSlNames(terms);
if (anyWild(terms))
list = wildAssociationFilter(termList, orLogic, col, conn, list);
else
list = tameAssociationFilter(termList, orLogic, col, conn, list);
}
return list;
}
struct subjInfo *stringAdvFilter(struct column *col,
struct sqlConnection *conn, struct subjInfo *list)
/* Do advanced filter on string in main table. */
{
char *wild = advFilterVal(col, "wild");
struct hash *keyHash = keyFileHash(col);
if (keyHash != NULL)
{
struct subjInfo *newList = NULL, *next, *si;
for (si = list; si != NULL; si = next)
{
char *cell = col->cellVal(col, si, conn);
next = si->next;
if (hashLookupUpperCase(keyHash, cell))
{
slAddHead(&newList, si);
}
freez(&cell);
}
slReverse(&newList);
list = newList;
}
if (wild != NULL)
{
boolean orLogic = advFilterOrLogic(col, "logic", TRUE);
struct subjInfo *newList = NULL, *next, *si;
struct slName *wildList = stringToSlNames(wild);
for (si = list; si != NULL; si = next)
{
char *cell = col->cellVal(col, si, conn);
next = si->next;
if (wildMatchList(cell, wildList, orLogic))
{
slAddHead(&newList, si);
}
freez(&cell);
}
slReverse(&newList);
list = newList;
}
hashFree(&keyHash);
return list;
}
void doPeakClusterListItemsAssayed()
/* Put up a page that shows all experiments associated with a cluster track. */
{
struct trackDb *clusterTdb = tdbForTableArg();
cartWebStart(cart, database, "List of items assayed in %s", clusterTdb->shortLabel);
struct sqlConnection *conn = hAllocConn(database);
char *inputTableFieldDisplay = trackDbSetting(clusterTdb, "inputTableFieldDisplay");
webPrintLinkTableStart();
if (inputTableFieldDisplay)
{
struct slName *fieldList = stringToSlNames(inputTableFieldDisplay);
printClusterTableHeader(fieldList, FALSE, FALSE, FALSE);
char *inputTrackTable = trackDbRequiredSetting(clusterTdb, "inputTrackTable");
printPeakClusterInfo(clusterTdb, cart, conn, inputTrackTable, fieldList, NULL);
}
else
errAbort("Missing required trackDb setting %s for track %s", "inputTableFieldDisplay",
clusterTdb->track);
webPrintLinkTableEnd();
hFreeConn(&conn);
}
static struct genePos *pfamAdvFilter(struct column *col,
struct sqlConnection *defaultConn, struct genePos *list)
/* Do advanced filter on for pfam. */
{
char *terms = advFilterVal(col, "terms");
if (terms != NULL)
{
struct sqlConnection *conn = sqlConnect(col->protDb);
char query[256];
struct sqlResult *sr;
struct dyString *dy = newDyString(1024);
char **row;
boolean orLogic = advFilterOrLogic(col, "logic", TRUE);
struct slName *term, *termList = stringToSlNames(terms);
struct hash *passHash = newHash(17);
struct hash *prevHash = NULL;
struct genePos *gp;
/* Build up hash of all genes. */
struct hash *geneHash = newHash(18);
for (gp = list; gp != NULL; gp = gp->next)
hashAdd(geneHash, gp->name, gp);
for (term = termList; term != NULL; term = term->next)
{
/* Build up a list of IDs of descriptions that match term. */
struct slName *idList = NULL, *id;
if (isPfamId(term->name))
{
idList = slNameNew(term->name);
}
else
{
char *sqlWild = sqlLikeFromWild(term->name);
sqlSafef(query, sizeof(query),
"select pfamAC from pfamDesc where description like '%s'",
sqlWild);
sr = sqlGetResult(conn, query);
while ((row = sqlNextRow(sr)) != NULL)
{
id = slNameNew(row[0]);
slAddHead(&idList, id);
}
sqlFreeResult(&sr);
}
if (idList != NULL)
{
/* Build up query that includes all IDs. */
dyStringClear(dy);
sqlDyStringPrintf(dy, "select name from %s where ", col->table);
sqlDyStringPrintf(dy, "value='%s'", idList->name);
for (id = idList->next; id != NULL; id = id->next)
sqlDyStringPrintf(dy, "or value='%s'", id->name);
/* Execute query and put matchers into hash. */
sr = sqlGetResult(defaultConn, dy->string);
while ((row = sqlNextRow(sr)) != NULL)
{
gp = hashFindVal(geneHash, row[0]);
if (gp != NULL)
{
char *name = gp->name;
if (prevHash == NULL || hashLookup(prevHash, name) != NULL)
hashStore(passHash, name);
}
}
sqlFreeResult(&sr);
slFreeList(&idList);
}
if (!orLogic)
{
hashFree(&prevHash);
if (term->next != NULL)
{
prevHash = passHash;
passHash = newHash(17);
}
}
}
list = weedUnlessInHash(list, passHash);
hashFree(&prevHash);
hashFree(&passHash);
dyStringFree(&dy);
sqlDisconnect(&conn);
}
return list;
}
void doFactorSource(struct sqlConnection *conn, struct trackDb *tdb, char *item, int start, int end)
/* Display detailed info about a cluster of TFBS peaks from other tracks. */
{
char extraWhere[256];
safef(extraWhere, sizeof extraWhere, "name='%s'", item);
int rowOffset;
struct sqlResult *sr = hRangeQuery(conn, tdb->table, seqName, start, end, extraWhere, &rowOffset);
char **row = sqlNextRow(sr);
struct factorSource *cluster = NULL;
if (row != NULL)
cluster = factorSourceLoad(row + rowOffset);
sqlFreeResult(&sr);
if (cluster == NULL)
errAbort("Error loading cluster from track %s", tdb->track);
char *sourceTable = trackDbRequiredSetting(tdb, "sourceTable");
char *factorLink = cluster->name;
char *vocab = trackDbSetting(tdb, "controlledVocabulary");
if (vocab != NULL)
{
char *file = cloneFirstWord(vocab);
factorLink = wgEncodeVocabLink(file, "term", factorLink, factorLink, factorLink, "");
}
printf("<B>Factor:</B> %s<BR>\n", factorLink);
printf("<B>Cluster Score (out of 1000):</B> %d<BR>\n", cluster->score);
printPos(cluster->chrom, cluster->chromStart, cluster->chromEnd, NULL, TRUE, item);
/* Get list of tracks we'll look through for input. */
char *inputTrackTable = trackDbRequiredSetting(tdb, "inputTrackTable");
char query[256];
sqlSafef(query, sizeof(query), "select tableName from %s where factor='%s' order by source",
inputTrackTable,
cluster->name);
/* Next do the lists of hits and misses. We have the hits from the non-zero signals in
* cluster->expScores. We need to figure out the sources actually assayed though
* some other way. We'll do this by one of two techniques. */
char *inputTableFieldDisplay = trackDbSetting(tdb, "inputTableFieldDisplay");
if (inputTableFieldDisplay != NULL)
{
struct slName *fieldList = stringToSlNames(inputTableFieldDisplay);
char *vocab = trackDbSetting(tdb, "controlledVocabulary");
/* In a new section put up list of hits. */
webNewSection("Assays for %s in Cluster", cluster->name);
webPrintLinkTableStart();
printClusterTableHeader(fieldList, TRUE, FALSE, TRUE);
printFactorSourceTableHits(cluster, conn, sourceTable,
inputTrackTable, fieldList, FALSE, vocab);
webPrintLinkTableEnd();
webNewSectionHeaderStart();
char sectionTitle[128];
safef(sectionTitle,
sizeof(sectionTitle),"Assays for %s Without Hits in Cluster", cluster->name);
jsBeginCollapsibleSectionOldStyle(cart, tdb->track, "cellNoHits", sectionTitle, FALSE);
webNewSectionHeaderEnd();
webPrintLinkTableStart();
printClusterTableHeader(fieldList, TRUE, FALSE, FALSE);
printFactorSourceTableHits(cluster, conn, sourceTable,
inputTrackTable, fieldList, TRUE, vocab);
webPrintLinkTableEnd();
jsEndCollapsibleSection();
}
else
{
errAbort("Missing required trackDb setting %s for track %s",
"inputTableFieldDisplay", tdb->track);
}
webNewSectionHeaderStart();
jsBeginCollapsibleSectionOldStyle(cart, tdb->track, "cellSources", "Cell Abbreviations", FALSE);
webNewSectionHeaderEnd();
hPrintFactorSourceAbbrevTable(conn, tdb);
jsEndCollapsibleSection();
doClusterMotifDetails(conn, tdb, cluster);
}
