void makeActiveImagePB(char *psOutput, char *psOutput2)
/* Make image and image map. */
{
char *mapName = "map";
int pixWidth, pixHeight;
char *answer;
char cond_str[255];
struct sqlConnection *conn;
struct sqlConnection *connCentral;
char query[256];
struct sqlResult *sr;
char **row;
int iypos;
char *blatGbDb;
char *sciName, *commonName;
char *spDisplayId;
char *oldDisplayId;
conn = sqlConnect(UNIPROT_DB_NAME);
hPrintf("<br><font size=4>Protein ");
hPrintf("<A HREF=\"http://www.uniprot.org/uniprot/%s\" TARGET=_blank><B>%s</B></A>\n",
proteinID, proteinID);
spDisplayId = spAccToId(conn, spFindAcc(conn, proteinID));
if (strstr(spDisplayId, spFindAcc(conn, proteinID)) == NULL)
{
hPrintf(" (aka %s", spDisplayId);
/* show once if the new and old displayId are the same */
oldDisplayId = oldSpDisplayId(spDisplayId);
if (oldDisplayId != NULL)
{
if (!sameWord(spDisplayId, oldDisplayId))
{
hPrintf(" or %s", oldSpDisplayId(spDisplayId));
}
}
hPrintf(")\n");
}
hPrintf(" %s\n", description);
hPrintf("</font><br>");
hPrintf("Organism: ");
/* get scientific and Genbank common name of this organism */
sciName = NULL;
commonName = NULL;
sqlSafefFrag(cond_str, sizeof(cond_str),"accession='%s'", proteinID);
answer = sqlGetField(PROTEOME_DB_NAME, "spXref3", "division", cond_str);
if (answer != NULL)
{
sqlSafefFrag(cond_str, sizeof(cond_str), "id=%s and nameType='scientific name'", answer);
sciName = sqlGetField(PROTEOME_DB_NAME, "taxonNames", "name", cond_str);
sqlSafefFrag(cond_str, sizeof(cond_str), "id=%s and nameType='genbank common name'", answer);
commonName = sqlGetField(PROTEOME_DB_NAME, "taxonNames", "name", cond_str);
}
if (sciName != NULL)
{
hPrintf("%s", sciName);
}
if (commonName != NULL)
{
hPrintf(" (%s)", commonName);
}
hPrintf("<br>");
protSeq = getAA(proteinID);
if (protSeq == NULL)
{
hUserAbort("%s is not a current valid entry in UniProtKB\n", proteinID);
}
protSeqLen = strlen(protSeq);
fflush(stdout);
iypos = 15;
doTracks(proteinID, mrnaID, protSeq, &iypos, psOutput);
if (!hTableExists(database, "pbStamp")) goto histDone;
pbScale = 3;
pixWidth = 765;
insideWidth = pixWidth-gfxBorder;
pixHeight = 350;
if (psOutput2)
{
vg2 = vgOpenPostScript(pixWidth, pixHeight, psOutput2);
}
else
{
trashDirFile(&gifTn2, "pbt", "pbt", ".png");
vg2 = vgOpenPng(pixWidth, pixHeight, gifTn2.forCgi, FALSE);
}
g_vg = vg2;
pbRed = vgFindColorIx(vg2, 0xf9, 0x51, 0x59);
pbBlue = vgFindColorIx(g_vg, 0x00, 0x00, 0xd0);
normalColor = pbBlue;
abnormalColor = pbRed;
bkgColor = vgFindColorIx(vg2, 255, 254, 232);
vgBox(vg2, 0, 0, insideWidth, pixHeight, bkgColor);
/* Start up client side map. */
mapName=cloneString("pbStamps");
hPrintf("\n<MAP Name=%s>\n", mapName);
vgSetClip(vg2, 0, gfxBorder, insideWidth, pixHeight - 2*gfxBorder);
iypos = 15;
/* Draw stamps. */
doStamps(proteinID, mrnaID, protSeq, vg2, &iypos);
/* Finish map. */
hPrintf("</MAP>\n");
/* Save out picture and tell html file about it. */
vgClose(&vg2);
hPrintf("<P>");
hPrintf("\n<IMG SRC=\"%s\" BORDER=1 WIDTH=%d HEIGHT=%d USEMAP=#%s><BR>",
gifTn2.forCgi, pixWidth, pixHeight, mapName);
if (proteinInSupportedGenome)
{
hPrintf("\n<A HREF=\"../goldenPath/help/pbTracksHelpFiles/pbTracksHelp.shtml#histograms\" TARGET=_blank>");
}
else
{
hPrintf("\n<A HREF=\"../goldenPath/help/pbTracksHelpFiles/pbTracksHelp.shtml#histograms\" TARGET=_blank>");
}
hPrintf("Explanation of Protein Property Histograms</A><BR>");
hPrintf("<P>");
histDone:
hPrintf("<P>");
fflush(stdout);
/* See if a UCSC Genome Browser exist for this organism. If so, display BLAT link. */
connCentral = hConnectCentral();
sqlSafef(query, sizeof(query),
"select defaultDb.name from dbDb, defaultDb where dbDb.scientificName='%s' and dbDb.name=defaultDb.name",
sciName);
sr = sqlGetResult(connCentral, query);
row = sqlNextRow(sr);
if (row != NULL)
{
blatGbDb = strdup(row[0]);
}
else
{
blatGbDb = NULL;
}
sqlFreeResult(&sr);
hDisconnectCentral(&connCentral);
if (proteinInSupportedGenome || (blatGbDb != NULL))
{
hPrintf("\n<B>UCSC Links:</B><BR>\n ");
hPrintf("<UL>\n");
/* Show GB links only if the protein belongs to a supported genome */
if (proteinInSupportedGenome)
{
doGenomeBrowserLink(proteinID, mrnaID, hgsidStr);
doGeneDetailsLink(proteinID, mrnaID, hgsidStr);
}
/* Show Gene Sorter link only if it is valid for this genome */
if (hgNearOk(database))
{
doGeneSorterLink(protDisplayID, mrnaID, hgsidStr);
}
/* Show BLAT link if we have UCSC Genome Browser for it */
if (blatGbDb != NULL)
{
doBlatLink(blatGbDb, sciName, commonName, protSeq);
}
hPrintf("</UL><P>");
}
/* This section shows various types of domains */
conn = sqlConnect(UNIPROT_DB_NAME);
domainsPrint(conn, proteinID);
hPrintf("<P>");
/* Do Pathway section only if the protein belongs to a supported genome */
if (proteinInSupportedGenome);
{
doPathwayLinks(proteinID, mrnaID);
}
printFASTA(proteinID, protSeq);
}
char *linkGetUrl(struct link *link, struct sqlConnection *conn,
char *geneId)
/* Return URL string if possible or NULL if not. FreeMem this when done. */
{
char query[512];
struct sqlResult *sr;
char **row;
char *url = NULL;
/* Some special case code here for things that need to
* do more than check a table. */
if (sameString(link->name, "family"))
{
if (!hgNearOk(database))
return NULL;
}
if (sameString(link->name, "protBrowser"))
{
if (!hgPbOk(database))
return NULL;
/* special processing for PB, since we need the protein ID, instead everything key off from gene ID */
/* use UniProt accession instead of displayID, because display ID sometimes changes */
if (swissProtAcc == NULL || swissProtAcc[0] == 0)
return NULL;
if (isRgdGene(conn))
{
safef(query, sizeof(query), "../cgi-bin/pbGlobal?proteinID=%s", swissProtAcc);
}
else
{
safef(query, sizeof(query), "../cgi-bin/pbTracks?db=%s&proteinID=%s", database, swissProtAcc);
}
return(cloneString(query));
}
if (sameString(link->name, "tbSchema"))
{
char *geneTable = genomeSetting("knownGene");
struct trackDb *tdb = hTrackDbForTrack(sqlGetDatabase(conn), geneTable);
struct dyString *dy = NULL;
if (tdb == NULL)
return NULL;
dy = newDyString(256);
dyStringPrintf(dy, link->url, tdb->grp, geneTable, geneTable);
trackDbFree(&tdb);
addLinkExtras(link, dy);
return dyStringCannibalize(&dy);
}
geneId = cloneAndCut(geneId, link->preCutAt);
safef(query, sizeof(query), link->idSql, geneId);
sr = sqlGetResult(conn, query);
row = sqlNextRow(sr);
if (row != NULL && row[0][0] != 0) /* If not null or empty */
{
struct dyString *dy = newDyString(0);
char *name = cloneAndCut(row[0], link->postCutAt);
dyStringPrintf(dy, link->url, name, row[1], row[2], row[3]);
addLinkExtras(link, dy);
url = dyStringCannibalize(&dy);
freez(&name);
}
sqlFreeResult(&sr);
freeMem(geneId);
return url;
}
void sortGenes(struct sqlConnection *conn)
/* Put up sort gene page. */
{
cartWebStart(cart, database, "Finding Candidate Genes for Gene Sorter");
if (!hgNearOk(database))
errAbort("Sorry, gene sorter not available for this database.");
/* Get list of regions. */
struct genoGraph *gg = ggFirstVisible();
double threshold = getThreshold();
struct bed3 *bed, *bedList = regionsOverThreshold(gg);
/* Figure out what table and column are the sorter's main gene set. */
struct hash *genomeRa = hgReadRa(genome, database, "hgNearData",
"genome.ra", NULL);
char *geneTable = hashMustFindVal(genomeRa, "geneTable");
char *idColumn = hashMustFindVal(genomeRa, "idColumn");
/* if marker labels were present when the file was uploaded, they are saved here */
char cgmName[256];
safef(cgmName, sizeof(cgmName), "%s.cgm", gg->binFileName);
struct lineFile *m = lineFileMayOpen(cgmName, TRUE);
char *cgmRow[4];
cgmRow[0] = ""; /* dummy row */
cgmRow[1] = "";
cgmRow[2] = "0";
cgmRow[3] = "0";
FILE *g = NULL;
int markerCount = 0;
struct tempName snpTn;
if (m)
{
/* Create custom column output file. */
trashDirFile(&snpTn, "hgg", "marker", ".mrk");
g = mustOpen(snpTn.forCgi, "w");
fprintf(g,
"column name=\"%s Markers\" shortLabel=\"%s Markers over threshold\" longLabel=\"%s Markers in regions over threshold\" "
"visibility=on priority=99 "
"\n"
, gg->shortLabel
, gg->shortLabel
, gg->shortLabel
);
}
/*** Build up hash of all transcriptHash that are in region. */
struct hash *transcriptHash = hashNew(16);
/* This loop handles one chromosome at a time. It depends on
* the bedList being sorted by chromosome. */
for (bed = bedList; bed != NULL; )
{
/* Make binKeeper and stuff in all regions in this chromosome into it. */
char *chrom = bed->chrom;
int chromSize = hChromSize(database, chrom);
struct binKeeper *bk = binKeeperNew(0, chromSize);
while (bed != NULL && sameString(chrom, bed->chrom))
{
binKeeperAdd(bk, bed->chromStart, bed->chromEnd, bed);
bed = bed->next;
}
struct binKeeper *bkGenes = NULL;
if (m)
bkGenes = binKeeperNew(0, chromSize);
/* Query database to find out bounds of all genes on this chromosome
* and if they overlap any of the regions then put them in the hash. */
char query[512];
safef(query, sizeof(query),
"select name,txStart,txEnd from %s where chrom='%s'", geneTable, chrom);
struct sqlResult *sr = sqlGetResult(conn, query);
char **row;
while ((row = sqlNextRow(sr)) != NULL)
{
char *name = row[0];
int start = sqlUnsigned(row[1]);
int end = sqlUnsigned(row[2]);
if (binKeeperAnyOverlap(bk, start, end))
{
hashStore(transcriptHash, name);
if (m)
binKeeperAdd(bkGenes, start, end, cloneString(name));
}
}
sqlFreeResult(&sr);
if (m)
{
/* Read cgm file if it exists, looking at all markers on this chromosome
* and if they overlap any of the regions and genes then output them. */
do
{
// marker, chrom, chromStart, val
char *marker = cgmRow[0];
char *chr = cgmRow[1];
int start = sqlUnsigned(cgmRow[2]);
int end = start+1;
double val = sqlDouble(cgmRow[3]);
int cmp = strcmp(chr,chrom);
if (cmp > 0)
break;
if (cmp == 0)
{
if (val >= threshold)
{
struct binElement *el, *bkList = binKeeperFind(bkGenes, start, end);
for (el = bkList; el; el=el->next)
{
/* output to custom column trash file */
fprintf(g, "%s %s\n", (char *)el->val, marker);
}
if (bkList)
{
++markerCount;
slFreeList(&bkList);
}
}
}
}
while (lineFileRow(m, cgmRow));
}
/* Clean up for this chromosome. */
binKeeperFree(&bk);
if (m)
{
/* For speed, we do not free up the values (cloned the kg names earlier) */
binKeeperFree(&bkGenes);
}
}
/* Get list of all transcripts in regions. */
struct hashEl *el, *list = hashElListHash(transcriptHash);
/* Create file with all matching gene IDs. */
struct tempName keyTn;
trashDirFile(&keyTn, "hgg", "key", ".key");
FILE *f = mustOpen(keyTn.forCgi, "w");
for (el = list; el != NULL; el = el->next)
fprintf(f, "%s\n", el->name);
carefulClose(&f);
/* Print out some info. */
hPrintf("Thresholding <i>%s</i> at %g. ", gg->shortLabel, threshold);
hPrintf("There are %d regions covering %lld bases.<BR>\n",
slCount(bedList), bedTotalSize((struct bed*)bedList) );
hPrintf("Installed a Gene Sorter filter that selects only genes in these regions.<BR>\n");
if (m)
{
hPrintf("There are %d markers in the regions over threshold that overlap knownGenes.<BR>\n", markerCount);
hPrintf("Installed a Gene Sorter custom column called \"%s Markers\" with these markers.<BR>\n", gg->shortLabel);
}
/* close custom column output file */
if (m)
{
lineFileClose(&m);
carefulClose(&g);
}
/* Stuff cart variable with name of file. */
char keyCartName[256];
safef(keyCartName, sizeof(keyCartName), "%s%s.keyFile",
advFilterPrefix, idColumn);
cartSetString(cart, keyCartName, keyTn.forCgi);
cartSetString(cart, customFileVarName, snpTn.forCgi);
char snpVisCartNameTemp[256];
char *snpVisCartName = NULL;
safef(snpVisCartNameTemp, sizeof(snpVisCartNameTemp), "%s%s Markers.vis",
colConfigPrefix, gg->shortLabel);
snpVisCartName = replaceChars(snpVisCartNameTemp, " ", "_");
cartSetString(cart, snpVisCartName, "1");
freeMem(snpVisCartName);
hPrintf("<FORM ACTION=\"../cgi-bin/hgNear\" METHOD=GET>\n");
cartSaveSession(cart);
hPrintf("<CENTER>");
cgiMakeButton("submit", "go to gene sorter");
hPrintf("</CENTER>");
hPrintf("</FORM>");
cartWebEnd();
}
