void writePcrResultTrack(struct gfPcrOutput *gpoList, char *db, char *target)
/* Write trash files and store their name in a cart variable. */
{
char *cartVar = pcrResultCartVar(db);
struct tempName bedTn, primerTn;
char buf[2048];
trashDirFile(&bedTn, "hgPcr", "hgPcr", ".psl");
trashDirFile(&primerTn, "hgPcr", "hgPcr", ".txt");
gfPcrOutputWriteAll(gpoList, "psl", NULL, bedTn.forCgi);
writePrimers(gpoList, primerTn.forCgi);
if (isNotEmpty(target))
safef(buf, sizeof(buf), "%s %s %s", bedTn.forCgi, primerTn.forCgi, target);
else
safef(buf, sizeof(buf), "%s %s", bedTn.forCgi, primerTn.forCgi);
cartSetString(cart, cartVar, buf);
}
static void h1n1DownloadPdb(char *item, char *pdbUrl, struct tempName *tmpPdb)
/* uncompress PDB to trash */
{
int inFd = netOpenHttpExt(pdbUrl, "GET", NULL);
int inFdRedir = 0;
char *pdbUrlRedir = NULL;
if (!netSkipHttpHeaderLinesHandlingRedirect(inFd, pdbUrl, &inFdRedir, &pdbUrlRedir))
errAbort("Unable to access predicted 3D structure file: %s", pdbUrl);
if (pdbUrlRedir != NULL)
{
close(inFd);
inFd = inFdRedir;
freez(&pdbUrlRedir);
}
trashDirFile(tmpPdb, "hgct", item, ".pdb");
FILE *outFh = mustOpen(tmpPdb->forCgi, "w");
struct lineFile *inLf = lineFileDecompressFd(pdbUrl, TRUE, inFd);
char *line;
while (lineFileNext(inLf, &line, NULL))
{
fputs(line, outFh);
fputc('\n', outFh);
}
lineFileClose(&inLf);
carefulClose(&outFh);
}
void doGetGreatOutput(void (*dispatch)())
{
struct tempName tn;
int saveStdout;
FILE *f;
trashDirFile(&tn, "great", cartSessionId(cart), ".bed");
f = fopen(tn.forCgi, "w");
/* We want to capture hgTables stdout output to a trash file
* which will later be fetched by Great via URL. */
/* Workaround because stdout stream is not assignable on some operating systems */
fflush(stdout);
saveStdout = dup(STDOUT_FILENO);
dup2(fileno(f),STDOUT_FILENO); /* closes STDOUT before setting it again */
fclose(f);
dispatch(); /* this writes to stdout */
/* restore stdout */
fflush(stdout);
dup2(saveStdout ,STDOUT_FILENO); /* closes STDOUT before setting it back to saved descriptor */
close(saveStdout);
cartRemove(cart, hgtaDoGreatOutput);
htmlOpen("Table Browser integration with GREAT");
doSubmitToGreat(tn.forCgi);
htmlClose();
}
char *userRegionsParse(char *db, char *regionsText, int maxRegions, int maxErrs,
int *retRegionCount, char **retWarnText)
/* Parse user regions entered as BED3, BED4 or chr:start-end optionally followed by name.
* Return name of trash file containing BED for parsed regions if regionsText contains
* valid regions; otherwise NULL.
* If maxRegions <= 0, it is ignored; likewise for maxErrs.
* If retRegionCount is non-NULL, it will be set to the number of valid parsed regions
* in the trash file.
* If retWarnText is non-NULL, it will be set to a string containing warning and error
* messages encountered while parsing input. */
{
char *trashFileName = NULL;
if (isNotEmpty(regionsText))
{
char *copy = cloneString(regionsText);
struct dyString *dyWarn = dyStringNew(0);
struct bed4 *bedList = parseRegionInput(db, copy, maxRegions, maxErrs, dyWarn);
if (retWarnText != NULL)
{
if (dyWarn->stringSize > 0)
*retWarnText = dyStringCannibalize(&dyWarn);
else
{
*retWarnText = NULL;
dyStringFree(&dyWarn);
}
}
int regionCount = slCount(bedList);
if (retRegionCount != NULL)
*retRegionCount = regionCount;
if (regionCount > 0)
{
struct tempName tn;
trashDirFile(&tn, "hgtData", "user", ".region");
trashFileName = cloneString(tn.forCgi);
FILE *f = mustOpen(trashFileName, "w");
struct bed4 *bed;
for (bed = bedList; bed; bed = bed->next )
{
char *name = bed->name ? bed->name : "";
fprintf(f, "%s\t%d\t%d\t%s\n",
bed->chrom, bed->chromStart, bed->chromEnd, name);
}
carefulClose(&f);
}
freeMem(copy);
}
else
{
if (retRegionCount != NULL)
*retRegionCount = 0;
if (retWarnText != NULL)
*retWarnText = NULL;
}
return trashFileName;
}
static boolean drawGtexRBoxplot(char *geneName, struct tissueSampleVals *tsvList,
boolean doLogTransform, char *version, struct tempName *pngTn)
/* Draw a box-and-whiskers plot from GTEx sample data, using R boxplot */
{
/* Create R data frame. This is a tab-sep file, one row per sample,
* with columns for sample, tissue, rpkm */
struct tempName dfTn;
trashDirFile(&dfTn, "hgc", "gtexGene", ".df.txt");
FILE *f = fopen(dfTn.forCgi, "w");
if (f == NULL)
errAbort("can't create temp file %s", dfTn.forCgi);
fprintf(f, "sample\ttissue\trpkm\n");
struct tissueSampleVals *tsv;
int sampleId=1;
int i;
for (tsv = tsvList; tsv != NULL; tsv = tsv->next)
{
int count = tsv->count;
// remove trailing parenthesized phrases as not worth label length
chopSuffixAt(tsv->description, '(');
for (i=0; i<count; i++)
fprintf(f, "%d\t%s\t%0.3f\n", sampleId++, tsv->description, tsv->vals[i]);
}
fclose(f);
// Plot to PNG file
if (!pngTn)
return FALSE;
trashDirFile(pngTn, "hgc", "gtexGene", ".png");
char cmd[256];
/* Exec R in quiet mode, without reading/saving environment or workspace */
safef(cmd, sizeof(cmd), "Rscript --vanilla --slave hgcData/gtexBoxplot.R %s %s %s %s %s %s",
geneName, dfTn.forCgi, pngTn->forHtml,
doLogTransform ? "log=TRUE" : "log=FALSE", "order=alpha", version);
//NOTE: use "order=score" to order bargraph by median RPKM, descending
int ret = system(cmd);
if (ret == 0)
return TRUE;
return FALSE;
}
static void getOutFile(char *pdbId, char *other,
struct tempName *outName)
/* generate output file names, other can be NULL*/
{
char baseName[PATH_LEN];
if (other != NULL)
safef(baseName, sizeof(baseName), "%s.%s", pdbId, other);
else
safecpy(baseName, sizeof(baseName), pdbId);
trashDirFile(outName, "lssnp", baseName, ".chimerax");
}
void handlePostscript()
/* Deal with Postscript output. */
{
struct tempName psTn;
struct tempName psTn2;
char *pdfFile = NULL;
trashDirFile(&psTn, "pbt", "pbt", ".eps");
trashDirFile(&psTn2, "pbt", "pbt2", ".eps");
printf("<H1>PostScript/PDF Output</H1>\n");
printf("PostScript images can be printed at high resolution "
"and edited by many drawing programs such as Adobe Illustrator.");
printf(" PDF can be viewed with Adobe Acrobat Reader.<BR><BR>\n");
doTrackForm(psTn.forCgi, psTn2.forCgi);
printf("<A HREF=\"%s\">Click here to download</A> "
"the current protein tracks graphic in PostScript. ", psTn.forCgi);
pdfFile = convertEpsToPdf(psTn.forCgi);
if (pdfFile != NULL)
{
printf("<BR><A HREF=\"%s\">Click here to download</A> "
"the current protein tracks graphic in PDF", pdfFile);
}
else
printf("<BR><BR>PDF format not available");
printf("<BR><BR><A HREF=\"%s\">Click here to download</A> "
"the current protein histograms graphic in PostScript. ", psTn2.forCgi);
pdfFile = convertEpsToPdf(psTn2.forCgi);
if (pdfFile != NULL)
{
printf("<BR><A HREF=\"%s\">Click here to download</A> "
"the current protein histograms graphic in PDF", pdfFile);
}
else
printf("<BR><BR>PDF format not available");
freez(&pdfFile);
}
static void getTrashFileNames(char *rsId, char epsFile[PATH_LEN], char pdfFile[PATH_LEN],
char pngFile[PATH_LEN])
/* Get stable trash file names (always in trash/hgc, based on geo->name so we can reuse). */
{
struct tempName tn;
trashDirFile(&tn, "hgc", "", "");
char trashDir[FILENAME_LEN];
splitPath(tn.forCgi, trashDir, NULL, NULL);
safef(epsFile, PATH_LEN, "%shgdpGeo_%s.eps", trashDir, rsId);
safef(pdfFile, PATH_LEN, "%shgdpGeo_%s.pdf", trashDir, rsId);
safef(pngFile, PATH_LEN, "%shgdpGeo_%s.png", trashDir, rsId);
}
static struct customTrack *beginCustomTrack(char *table, int fields,
boolean doCt, boolean doWigHdr, boolean doDataPoints)
/* If doCt, return a new custom track object from TB cart settings and params;
* if !doCt, return NULL but print out a header line. */
{
char *ctName = cgiUsualString(hgtaCtName, table);
char *ctDesc = cgiUsualString(hgtaCtDesc, table);
char *ctUrl = cgiUsualString(hgtaCtUrl, "");
char *ctVis = cgiUsualString(hgtaCtVis, "dense");
int visNum = (int) hTvFromStringNoAbort(ctVis);
struct customTrack *ctNew = NULL;
if (visNum < 0)
visNum = 0;
if (doCt)
{
ctNew = newCt(ctName, ctDesc, visNum, ctUrl, fields);
if (doDataPoints)
{
struct dyString *wigSettings = newDyString(0);
struct tempName tn;
trashDirFile(&tn, "ct", hgtaCtTempNamePrefix, ".wib");
ctNew->wibFile = cloneString(tn.forCgi);
char *wiggleFile = cloneString(ctNew->wibFile);
chopSuffix(wiggleFile);
strcat(wiggleFile, ".wia");
ctNew->wigAscii = cloneString(wiggleFile);
chopSuffix(wiggleFile);
/* .wig file will be created upon encoding in customFactory */
/*strcat(wiggleFile, ".wig");
ctNew->wigFile = cloneString(wiggleFile);
*/
ctNew->wiggle = TRUE;
dyStringPrintf(wigSettings,
"type wiggle_0\nwibFile %s\n", ctNew->wibFile);
ctNew->tdb->settings = dyStringCannibalize(&wigSettings);
freeMem(wiggleFile);
}
}
else
{
if (doWigHdr)
hPrintf("track type=wiggle_0 name=\"%s\" description=\"%s\" "
"visibility=%d url=%s \n",
ctName, ctDesc, visNum, ctUrl);
else
hPrintf("track name=\"%s\" description=\"%s\" visibility=%d url=%s \n",
ctName, ctDesc, visNum, ctUrl);
}
return ctNew;
}
void doDocIdMeta(struct cart *theCart)
{
char *docId = cartString(theCart, "docId");
cartWebStart(cart, database, "ENCODE DCC: Metadata for docId %s",docId);
struct sqlConnection *conn = sqlConnect(database);
char query[10 * 1024];
struct sqlResult *sr;
char **row;
struct tempName tn;
trashDirFile(&tn, "docId", "meta", ".txt");
char *tempFile = tn.forCgi;
boolean beenHere = FALSE;
printf("<a href=docIdView?db=%s> Return </a><BR>", database);
safef(query, sizeof query, "select * from %s where ix=%s", docIdTable,docId);
sr = sqlGetResult(conn, query);
while ((row = sqlNextRow(sr)) != NULL)
{
if (beenHere)
errAbort("found more than one record with ix value %s in table %s",
docId, docIdTable);
beenHere = TRUE;
struct docIdSub *docIdSub = docIdSubLoad(row);
cgiDecode(docIdSub->metaData, docIdSub->metaData, strlen(docIdSub->metaData));
//printf("tempFile is %s\n<BR>", tempFile);
FILE *f = mustOpen(tempFile, "w");
fwrite(docIdSub->metaData, strlen(docIdSub->metaData), 1, f);
fclose(f);
boolean validated;
struct mdbObj *mdbObj = mdbObjsLoadFromFormattedFile(tempFile, &validated);
unlink(tempFile);
if (mdbObj->next)
errAbort("got more than one metaObj from metaData");
for(; mdbObj; mdbObj = mdbObj->next)
{
printf("<BR>object: %s\n", mdbObj->obj);
struct mdbVar *vars = mdbObj->vars;
for(; vars; vars = vars->next)
{
printf("<BR>%s %s\n", vars->var, vars->val);
}
}
}
cartWebEnd();
}
void lsSnpPdbChimeraGenericLink(char *pdbSpec, char *script,
char *trashDirName, char *trashBaseName,
struct tempName *chimerax)
/* Generate a chimerax file for the given pdbSpec, which can be a PDB id or a
* URL. Copies in the lsSnpPdbChimera.py file and then adds optional script python code.
* Fills in chimerax structure.
* FIXME: This is an experiment for H1N1 flu browser, this function has
* nothing to do with LS/SNP. If we decide to keep this, this should be
* split into a generic chimera module.
*/
{
trashDirFile(chimerax, trashDirName, trashBaseName, ".chimerax");
FILE *xfh = chimeraxBegin(chimerax->forCgi);
fprintf(xfh, "\ndisplayPdb(\"%s\", ())\n", pdbSpec);
if (script != NULL)
fputs(script, xfh);
chimeraxEnd(&xfh);
}
static void mkH1n1StructData(char *gene, char *idPairFile, char *highlightId,
struct tempName *imageFile, struct tempName *chimeraScript)
/* generate 3D structure files; difference highlighting is generate only idPairFile or
* idPairFile, if specified, but not both. */
{
struct tempName prefix;
trashDirFile(&prefix, "hgct", gene, "tmp");
char idFile[PATH_LEN], idArg[PATH_LEN], logFile[PATH_LEN], cmd[2*PATH_LEN];
idArg[0] = '\0';
if ((idPairFile != NULL) || (highlightId != NULL))
{
safef(idFile, sizeof(idFile), "%s.ids", prefix.forCgi);
if (idPairFile != NULL)
{
// extract first column
safef(cmd, sizeof(cmd), "cut -f 1 %s >%s", idPairFile, idFile);
if (system(cmd) != 0)
errAbort("extracting protein ids failed: %s", cmd);
}
else
{
FILE *fh = mustOpen(idFile, "w");
fprintf(fh, "%s\n", highlightId);
carefulClose(&fh);
}
safef(idArg, sizeof(idArg), "--ids %s", idFile);
}
// dynamic_highlight.pl knows locations of model files
safef(logFile, sizeof(logFile), "%s.log", prefix.forCgi);
safef(cmd, sizeof(cmd), "%s/dynamic_highlight.pl --rasmol --chimera --protein %s --consensus 0602 %s --base %s >%s 2>&1",
getH1n1StructDir(), gene, idArg, prefix.forCgi, logFile);
if (system(cmd) != 0)
errAbort("creation of 3D structure highlight files failed: %s", cmd);
// output names are all predefined by script relative to prefix
tempNameFromPrefix(imageFile, &prefix, "_highlight.jpg");
tempNameFromPrefix(chimeraScript, &prefix, "_highlight.cmd");
}
void doPastedIdentifiers(struct sqlConnection *conn)
/* Process submit in paste identifiers page. */
{
char *idText = trimSpaces(cartString(cart, hgtaPastedIdentifiers));
htmlOpen("Table Browser (Input Identifiers)");
if (isNotEmpty(idText))
{
/* Write terms to temp file, checking whether they have matches, and
* save temp file name. */
boolean saveIdText = (strlen(idText) < MAX_IDTEXT);
char *idTextForLf = saveIdText ? cloneString(idText) : idText;
struct lineFile *lf = lineFileOnString("idText", TRUE, idTextForLf);
char *line, *word;
struct tempName tn;
FILE *f;
int totalTerms = 0, foundTerms = 0;
struct slName* missingTerms = NULL;
struct dyString *exampleMissingIds = dyStringNew(256);
char *actualDb = database;
if (sameWord(curTable, WIKI_TRACK_TABLE))
actualDb = wikiDbName();
struct hTableInfo *hti = maybeGetHti(actualDb, curTable, conn);
char *idField = getIdField(actualDb, curTrack, curTable, hti);
if (idField == NULL)
{
warn("Sorry, I can't tell which field of table %s to treat as the "
"identifier field.", curTable);
webNewSection("Table Browser");
cartRemove(cart, hgtaIdentifierDb);
cartRemove(cart, hgtaIdentifierTable);
cartRemove(cart, hgtaIdentifierFile);
mainPageAfterOpen(conn);
htmlClose();
return;
}
struct slName *allTerms = NULL, *term;
while (lineFileNext(lf, &line, NULL))
{
while ((word = nextWord(&line)) != NULL)
{
term = slNameNew(word);
slAddHead(&allTerms, term);
totalTerms++;
}
}
slReverse(&allTerms);
lineFileClose(&lf);
char *extraWhere = NULL;
int maxIdsInWhere = cartUsualInt(cart, "hgt_maxIdsInWhere", DEFAULT_MAX_IDS_IN_WHERE);
if (totalTerms > 0 && totalTerms <= maxIdsInWhere)
extraWhere = slNameToInExpression(idField, allTerms);
struct lm *lm = lmInit(0);
struct hash *matchHash = getAllPossibleIds(conn, lm, idField, extraWhere);
trashDirFile(&tn, "hgtData", "identifiers", ".key");
f = mustOpen(tn.forCgi, "w");
for (term = allTerms; term != NULL; term = term->next)
{
struct slName *matchList = NULL, *match;
if (matchHash == NULL)
{
matchList = slNameNew(term->name);
}
else
{
/* Support multiple alias->id mappings: */
char upcased[1024];
safecpy(upcased, sizeof(upcased), term->name);
touppers(upcased);
struct hashEl *hel = hashLookup(matchHash, upcased);
if (hel != NULL)
{
matchList = slNameNew((char *)hel->val);
while ((hel = hashLookupNext(hel)) != NULL)
{
match = slNameNew((char *)hel->val);
slAddHead(&matchList, match);
}
}
}
if (matchList != NULL)
{
foundTerms++;
for (match = matchList; match != NULL; match = match->next)
{
mustWrite(f, match->name, strlen(match->name));
mustWrite(f, "\n", 1);
}
}
else
{
slAddHead(&missingTerms, slNameNew(term->name));
}
}
slReverse(&missingTerms);
carefulClose(&f);
cartSetString(cart, hgtaIdentifierDb, database);
cartSetString(cart, hgtaIdentifierTable, curTable);
cartSetString(cart, hgtaIdentifierFile, tn.forCgi);
if (saveIdText)
freez(&idTextForLf);
else
cartRemove(cart, hgtaPastedIdentifiers);
int missingCount = totalTerms - foundTerms;
if (missingCount > 0)
{
char *xrefTable, *aliasField;
getXrefInfo(conn, &xrefTable, NULL, &aliasField);
boolean xrefIsSame = xrefTable && sameString(curTable, xrefTable);
struct tempName tn;
trashDirFile(&tn, "hgt/missingIds", cartSessionId(cart), ".tmp");
FILE *f = mustOpen(tn.forCgi, "w");
int exampleCount = 0;
for (term = missingTerms; term != NULL; term = term->next)
{
if (exampleCount < 10)
{
++exampleCount;
dyStringPrintf(exampleMissingIds, "%s\n", term->name);
}
fprintf(f, "%s\n", term->name);
}
carefulClose(&f);
dyStringPrintf(exampleMissingIds, "\n<a href=%s>Complete list of missing identifiers<a>\n", tn.forHtml);
warn("Note: %d of the %d given identifiers have no match in "
"table %s, field %s%s%s%s%s. "
"Try the \"describe table schema\" button for more "
"information about the table and field.\n"
"%d %smissing identifier(s):\n"
"%s\n",
(totalTerms - foundTerms), totalTerms,
curTable, idField,
(xrefTable ? (xrefIsSame ? "" : " or in alias table ") : ""),
(xrefTable ? (xrefIsSame ? "" : xrefTable) : ""),
(xrefTable ? (xrefIsSame ? " or in field " : ", field ") : ""),
(xrefTable ? aliasField : ""),
exampleCount,
exampleCount < missingCount ? "example " : "",
exampleMissingIds->string
);
webNewSection("Table Browser");
}
lmCleanup(&lm);
hashFree(&matchHash);
}
else
{
cartRemove(cart, hgtaIdentifierFile);
}
mainPageAfterOpen(conn);
htmlClose();
}
static struct labeledFile *parseToLabeledFiles(struct customPp *cpp,
int colCount, char *formatType, char *markerType,
boolean firstLineLabels, struct sqlConnection *conn, boolean report)
/* Parse out cpp until next track, creating a list of labeled
* binary files. */
{
/* Allocate a labeledFile for each column of real data. */
struct labeledFile *fileList = NULL, *fileEl;
int posColCount = markerCols(markerType);
int i;
for (i=posColCount; i<colCount; ++i)
{
struct tempName tempName;
char buf[16];
safef(buf, sizeof(buf), "hggUp%d", i);
trashDirFile(&tempName, "hgg", buf, ".cgb");
safef(buf, sizeof(buf), "%d", i+1-posColCount);
fileEl = labeledFileNew(tempName.forCgi, buf);
slAddHead(&fileList, fileEl);
}
slReverse(&fileList);
boolean ok = FALSE;
if (sameString(markerType, cgfMarkerGenomic))
ok = mayProcessGenomic(conn, cpp, colCount, formatType,
firstLineLabels, fileList, report);
else if (sameString(markerType, cgfMarkerSts))
ok = mayProcessDb(conn, cpp, colCount, formatType,
firstLineLabels, fileList, "stsMap",
"select chrom,round((chromStart+chromEnd)*0.5),name from %s",
"stsAlias", "select alias,trueName from %s", report, FALSE);
else if (sameString(markerType, cgfMarkerSnp))
{
char *snpTable = hFindLatestSnpTableConn(conn, NULL);
if (snpTable == NULL)
errAbort("No SNP table in %s", sqlGetDatabase(conn));
char *query = "select chrom,chromStart,name from %s";
ok = mayProcessDb(conn, cpp, colCount, formatType,
firstLineLabels, fileList, snpTable,
query, NULL, NULL, report, TRUE);
}
else if (sameString(markerType, cgfMarkerAffy100))
{
warn("Support for Affy 100k chip coming soon.");
}
else if (sameString(markerType, cgfMarkerAffy500)
|| sameString(markerType, cgfMarkerAffy6)
|| sameString(markerType, cgfMarkerAffy6SV)
|| sameString(markerType, cgfMarkerHumanHap300)
|| sameString(markerType, cgfMarkerHumanHap550)
|| sameString(markerType, cgfMarkerHumanHap650)
|| sameString(markerType, cgfMarkerHumanHap1M)
|| sameString(markerType, cgfMarkerAgilentCgh244A)
)
{
char *table = "";
if (sameString(markerType, cgfMarkerAffy500)) table = affy500Table;
if (sameString(markerType, cgfMarkerAffy6)) table = affy6Table;
if (sameString(markerType, cgfMarkerAffy6SV)) table = affy6SVTable;
if (sameString(markerType, cgfMarkerHumanHap300)) table = illumina300Table;
if (sameString(markerType, cgfMarkerHumanHap550)) table = illumina550Table;
if (sameString(markerType, cgfMarkerHumanHap650)) table = illumina650Table;
if (sameString(markerType, cgfMarkerHumanHap1M)) table = illumina1MTable;
if (sameString(markerType, cgfMarkerAgilentCgh244A)) table = agilentCgh244ATable;
if (!sqlTableExists(conn, table))
errAbort("Sorry, no data for %s on this assembly.",
markerType);
ok = mayProcessDb(conn, cpp, colCount, formatType,
firstLineLabels, fileList, table,
"select chrom,chromStart,name from %s", NULL, NULL, report, FALSE);
}
else
{
errAbort("Unknown identifier format. markerType=%s", markerType);
}
if (ok)
return fileList;
else
{
noWarnAbort();
return NULL;
}
}
void doStandard(struct cart *theCart)
{
cart = theCart;
cartWebStart(cart, database, "ENCODE DCC Submissions");
struct sqlConnection *conn = sqlConnect(database);
struct docIdSub *docIdSub;
char query[10 * 1024];
struct sqlResult *sr;
char **row;
struct tempName tn;
trashDirFile(&tn, "docId", "meta", ".txt");
char *tempFile = tn.forCgi;
//printf("tempFile is %s\n<BR>", tempFile);
// <Data type> <Cell Type> <Key Metadata> <View>
printf("<table border=1><tr>");
printf("<th>dataType</th>");
printf("<th>cell type</th>");
printf("<th>metadata</th>");
printf("<th>view</th>");
printf("<th>fileType</th>");
printf("<th>file</th>");
printf("<th>lab</th>");
printf("<th>assembly</th>");
printf("<th>subId</th>");
printf("<th>val-report</th>");
printf("</tr>\n");
safef(query, sizeof query, "select * from %s", docIdTable);
sr = sqlGetResult(conn, query);
while ((row = sqlNextRow(sr)) != NULL)
{
docIdSub = docIdSubLoad(row);
verbose(2, "ix %d\n", docIdSub->ix);
verbose(2, "submitDate %s\n", docIdSub->submitDate);
verbose(2, "md5sum %s\n", docIdSub->md5sum);
verbose(2, "valReport %s\n", docIdSub->valReport);
verbose(2, "metaData %s\n", docIdSub->metaData);
verbose(2, "submitPath %s\n", docIdSub->submitPath);
verbose(2, "submitter %s\n", docIdSub->submitter);
cgiDecode(docIdSub->metaData, docIdSub->metaData, strlen(docIdSub->metaData));
//printf("tempFile %s\n", tempFile);
FILE *f = mustOpen(tempFile, "w");
fwrite(docIdSub->metaData, strlen(docIdSub->metaData), 1, f);
fclose(f);
boolean validated;
struct mdbObj *mdbObj = mdbObjsLoadFromFormattedFile(tempFile, &validated);
unlink(tempFile);
// <Data type> <Cell Type> <Key Metadata> <View>
char *docIdType = mdbObjFindValue(mdbObj, "type");
char *docIdComposite = mdbObjFindValue(mdbObj, "composite");
char buffer[10 * 1024];
safef(buffer, sizeof buffer, "%d", docIdSub->ix);
if (sameString(database, "encpipeline_beta"))
docIdDir = docIdDirBeta;
printf("<tr>");
printf("<td>%s</td> ", mdbObjFindValue(mdbObj, "dataType"));
printf("<td>%s</td> ", mdbObjFindValue(mdbObj, "cell"));
struct dyString *str = newDyString(100);
addValue(str, mdbObjFindValue(mdbObj, "antibody"));
addValue(str, mdbObjFindValue(mdbObj, "treatment"));
addValue(str, mdbObjFindValue(mdbObj, "rnaExtract"));
addValue(str, mdbObjFindValue(mdbObj, "localization"));
printf("<td>%s<a href=docIdView?docId=%s&db=%s&meta=\"\"> ...</a></td>", str->string,buffer, database);
freeDyString(&str);
printf("<td>%s</td> ", mdbObjFindValue(mdbObj, "view"));
printf("<td>%s</td> ", mdbObjFindValue(mdbObj, "type"));
printf("<td><a href=%s> %s</a></td>",
docIdGetPath(buffer, docIdDir, docIdType, NULL) ,
docIdDecorate(docIdComposite,docIdSub->ix));
char *lab = mdbObjFindValue(mdbObj, "lab");
char *subId = mdbObjFindValue(mdbObj, "subId");
printf("<td><a href=docIdView?docId=%s&db=%s&lab=\"%s\"> %s</a></td>",buffer, database, subId, lab);
printf("<td>%s</td> ", mdbObjFindValue(mdbObj, "assembly"));
printf("<td>%s</td> ", subId);
printf("<td><a href=docIdView?docId=%s&db=%s&report=\"\"> report</a></td>", buffer, database);
printf("</tr>\n");
}
printf("</table>");
sqlFreeResult(&sr);
sqlDisconnect(&conn);
cartWebEnd();
}
void saveSubjList(struct subjInfo *subjListIn)
/* save the filtered list of subject gisaids to a file for other applications to use */
{
struct subjInfo *subjList;
char *outName = cartOptionalString(cart, gisaidSubjList);
char *outName2= cartOptionalString(cart, gisaidSeqList);
char *outName3= cartOptionalString(cart, gisaidAaSeqList);
char *outNameTemp;
struct tempName tn;
struct tempName tn2;
struct tempName tn3;
struct tempName tnTemp;
struct sqlResult *sr;
char **row;
char query[255];
int cnt;
char cmd[512];
if (!outName)
{
trashDirFile(&tn, "ct", "gisaidSubj", ".list");
outName = tn.forCgi;
}
if (!outName2)
{
trashDirFile(&tn2, "ct", "gisaidSeq", ".list");
outName2 = tn2.forCgi;
}
if (!outName3)
{
trashDirFile(&tn3, "ct", "gisaidAaSeq", ".list");
outName3 = tn3.forCgi;
}
trashDirFile(&tnTemp, "ct", "gisaidListTemp", ".list");
outNameTemp = tnTemp.forCgi;
FILE *outF = mustOpen(outName,"w");
FILE *outF2= mustOpen(outName2,"w");
FILE *outF3= mustOpen(outName3,"w");
cnt = 0;
subjList = subjListIn;
while (subjList)
{
fprintf(outF, "%s\n", subjList->fields[1]);
sqlSafef(query, sizeof(query),
"select distinct seqId from h1n1SeqXref where islId='%s'",
subjList->fields[0]);
sr = sqlGetResult(conn, query);
while ((row = sqlNextRow(sr)) != NULL)
{
/* Remove "ss." from the front of the DNA sequence ID,
so that they could be used both for DNA and protein MSA maf display */
fprintf(outF2, "%s\t%s\n", row[0], subjList->fields[1]);
cnt++;
}
sqlFreeResult(&sr);
subjList=subjList->next;
}
subjList = subjListIn;
while (subjList)
{
fprintf(outF, "%s\n", subjList->fields[0]);
sqlSafef(query, sizeof(query),
"select distinct seqId, geneSymbol from h1n1SeqXref where islId='%s'",
subjList->fields[0]);
sr = sqlGetResult(conn, query);
while ((row = sqlNextRow(sr)) != NULL)
{
/* Remove "ss." from the front of the DNA sequence ID,
so that they could be used both for DNA and protein MSA maf display */
fprintf(outF3, "%s_%s\t%s\n", row[0], row[1], subjList->fields[1]);
cnt++;
}
sqlFreeResult(&sr);
subjList=subjList->next;
}
carefulClose(&outF);
carefulClose(&outF2);
carefulClose(&outF3);
/* sort -u to make the selection lists unique */
sprintf(cmd, "cat %s |sort -u >%s", outName, outNameTemp);
system(cmd);
sprintf(cmd, "mv %s %s", outNameTemp, outName);
system(cmd);
sprintf(cmd, "cat %s |sort -u >%s", outName2, outNameTemp);
system(cmd);
sprintf(cmd, "mv %s %s", outNameTemp, outName2);
system(cmd);
sprintf(cmd, "cat %s |sort -u >%s", outName3, outNameTemp);
system(cmd);
sprintf(cmd, "mv %s %s", outNameTemp, outName3);
system(cmd);
cartSetString(cart, gisaidSubjList, outName);
cartSetString(cart, gisaidSeqList, outName2);
cartSetString(cart, gisaidAaSeqList, outName3);
}
void doSubmitUserRegions(struct sqlConnection *conn)
/* Process submit in set regions page. */
{
char *idText = trimSpaces(cartString(cart, hgtaEnteredUserRegions));
char *userRegionFile = trimSpaces(cartString(cart, hgtaEnteredUserRegionFile));
boolean hasData = (idText != NULL && idText[0] != 0) ||
(userRegionFile != NULL && userRegionFile[0] != 0);
/* beware, the string pointers from cartString() point to strings in the
* cart hash. If they are manipulated and changed, they will get saved
* back to the cart in their changed form. You don't want to be
* altering them like that. Thus, the idText is duplicated below with
* the cloneString(idText)
*/
htmlOpen("Table Browser (Region definitions)");
/* presence of fileName text overrides previously existing text area
* contents
*/
if (userRegionFile != NULL && userRegionFile[0] != 0)
{
idText = cloneString(userRegionFile);
cartRemove(cart, hgtaEnteredUserRegions);
cartRemove(cart, hgtaUserRegionsFile);
cartSetString(cart, hgtaEnteredUserRegions, idText);
}
else
idText = cloneString(idText);
char *lineLimitText = limitText(idText);
if ( (strlen(lineLimitText) > 0) && (strlen(lineLimitText) != strlen(idText)) )
{
freeMem(idText);
idText = lineLimitText;
cartSetString(cart, hgtaEnteredUserRegions, lineLimitText);
}
else
freeMem(lineLimitText);
if (hasData)
{
struct tempName tn;
FILE *f;
struct bed *bedEl;
struct bed *bedList = parseRegionInput(idText);
if (NULL == bedList)
errAbort("no valid data points found in input");
trashDirFile(&tn, "hgtData", "user", ".region");
f = mustOpen(tn.forCgi, "w");
for (bedEl = bedList; bedEl; bedEl = bedEl->next )
{
if (bedEl->name)
fprintf(f, "%s\t%d\t%d\t%s\n",
bedEl->chrom, bedEl->chromStart, bedEl->chromEnd, bedEl->name);
else
fprintf(f, "%s\t%d\t%d\n",
bedEl->chrom, bedEl->chromStart, bedEl->chromEnd);
}
carefulClose(&f);
cartSetString(cart, hgtaUserRegionsDb, database);
cartSetString(cart, hgtaUserRegionsTable, curTable);
cartSetString(cart, hgtaUserRegionsFile, tn.forCgi);
cartSetString(cart, hgtaRegionType, hgtaRegionTypeUserRegions);
if (strlen(idText) > 64 * 1024)
cartRemove(cart, hgtaEnteredUserRegions);
}
else
{
cartRemove(cart, hgtaUserRegionsFile);
cartRemove(cart, hgtaEnteredUserRegionFile);
cartRemove(cart, hgtaRegionType);
}
mainPageAfterOpen(conn);
htmlClose();
}
void doFrame(struct sqlConnection *conn, boolean forceImageToList)
/* Make a html frame page. Fill frame with thumbnail, control bar,
* and image panes. */
{
int imageId = cartUsualInt(cart, hgpId, 0);
char *sidUrl = cartSidUrlString(cart);
char *listSpec = cartUsualString(cart, hgpListSpec, "");
struct tempName matchTempName;
char *matchFile = NULL;
struct visiMatch *matchList = visiSearch(conn, listSpec);
#ifdef SOON
if (!cartUsualBoolean(cart, hgpIncludeMutants, FALSE))
matchList = removeMutants(conn, matchList);
#endif /* SOON */
matchList = onePerImageFile(conn, matchList);
weighMatches(conn, matchList);
slSort(&matchList, visiMatchCmpWeight);
if (forceImageToList)
{
if (matchList != NULL)
imageId = matchList->imageId;
else
imageId = 0;
}
trashDirFile(&matchTempName, "vg", "visiMatch", ".tab");
matchFile = matchTempName.forCgi;
saveMatchFile(matchFile, matchList);
cartSetString(cart, hgpMatchFile, matchFile);
cartSetInt(cart, hgpId, imageId);
//puts("\n");
puts("<!DOCTYPE HTML PUBLIC \"-//W3C//DTD HTML 4.01 Frameset//EN\" \"http://www.w3.org/TR/html4/frameset.dtd\">");
printf("<HTML>\n");
printf("<HEAD>\n");
printf("<TITLE>\n");
printf("%s ", hgVisiGeneShortName());
printf("%s",titleMessage);
printf("</TITLE>\n");
printf("</HEAD>\n");
printf(" <frameset rows=\"27,*\">\n");
printf(" <frame name=\"controls\" src=\"%s?%s=go&%s&%s=%d\" noresize marginwidth=\"0\" marginheight=\"0\" frameborder=\"0\">\n",
hgVisiGeneCgiName(), hgpDoControls, sidUrl, hgpId, imageId);
printf(" <frameset cols=\"230,*\"> \n");
printf(" <frame src=\"%s?%s=go&%s&%s=%d\" noresize frameborder=\"0\" name=\"list\">\n",
hgVisiGeneCgiName(), hgpDoThumbnails, sidUrl, hgpId, imageId);
printf(" <frame src=\"%s?%s=go&%s&%s=%d\" name=\"image\" noresize frameborder=\"0\">\n",
hgVisiGeneCgiName(), hgpDoImage, sidUrl, hgpId, imageId);
printf(" </frameset>\n");
printf(" <noframes>\n");
printf(" <body>\n");
printf(" <p>This web page uses frames, but your browser doesn't support them.</p>\n");
printf(" </body>\n");
printf(" </noframes>\n");
printf("</frameset>\n");
printf("</HTML>\n");
}
void saveSubjList(struct subjInfo *subjList)
/* save the filtered list of subject gsids to a file for other applications to use */
{
char *outName = cartOptionalString(cart, gsidSubjList);
char *outName2= cartOptionalString(cart, gsidSeqList);
struct tempName tn;
struct tempName tn2;
struct sqlResult *sr;
char **row;
char query[255];
char *chp;
int cnt;
if (!outName)
{
trashDirFile(&tn, "ct", "gsidSubj", ".list");
outName = tn.forCgi;
}
if (!outName2)
{
trashDirFile(&tn2, "ct", "gsidSeq", ".list");
outName2 = tn2.forCgi;
}
FILE *outF = mustOpen(outName,"w");
FILE *outF2= mustOpen(outName2,"w");
cnt = 0;
while (subjList)
{
fprintf(outF, "%s\n", subjList->fields[0]);
sqlSafef(query, sizeof(query),
"select dnaSeqId from gsIdXref where subjId='%s'",
subjList->fields[0]);
sr = sqlGetResult(conn, query);
while ((row = sqlNextRow(sr)) != NULL)
{
/* Remove "ss." from the front of the DNA sequence ID,
so that they could be used both for DNA and protein MSA maf display */
chp = strstr(row[0], "ss.");
if (chp != NULL)
{
fprintf(outF2, "%s\t%s\n", chp+3L, subjList->fields[0]);
}
else
{
fprintf(outF2, "%s\t%s\n", row[0], subjList->fields[0]);
}
cnt++;
}
sqlFreeResult(&sr);
subjList=subjList->next;
}
carefulClose(&outF);
carefulClose(&outF2);
cartSetString(cart, gsidSubjList, outName);
cartSetString(cart, gsidSeqList, outName2);
}
void doTracks(char *proteinID, char *mrnaID, char *aa, int *yOffp, char *psOutput)
/* draw various protein tracks */
{
int l;
char aaOrigOffsetStr[20];
int hasResFreq;
char uniProtDbName[50];
char *protDbDate;
char *chrom;
char strand;
char *kgId, *kgPep, *protPep;
char cond_str[255];
char *answer;
//int i, ll;
//char *chp1, *chp2;
g_font = mgSmallFont();
safef(pbScaleStr, sizeof(pbScaleStr), "%d", pbScale);
if (psOutput != NULL)
{
pbScale = atoi(cartOptionalString(cart, "pbt.pbScaleStr"));
}
if (cgiOptionalString("trackOffset") != NULL)
{
trackOrigOffset = atoi(cgiOptionalString("trackOffset"));
}
if (cgiOptionalString("pbScaleStr") != NULL)
{
pbScale = atoi(cgiOptionalString("pbScaleStr"));
}
if (cgiOptionalString("pbScale") != NULL)
{
scaleButtonPushed = TRUE;
if (strcmp(cgiOptionalString("pbScale"), "1/6") == 0) pbScale = 1;
if (strcmp(cgiOptionalString("pbScale"), "1/2") == 0) pbScale = 3;
if (strcmp(cgiOptionalString("pbScale"), "FULL") == 0) pbScale = 6;
if (strcmp(cgiOptionalString("pbScale"), "DNA") == 0) pbScale =22;
safef(pbScaleStr, sizeof(pbScaleStr), "%d", pbScale);
cgiMakeHiddenVar("pbScaleStr", pbScaleStr);
}
else
{
scaleButtonPushed = FALSE;
}
if (psOutput == NULL)
{
if (cgiVarExists("pbt.left3"))
{
relativeScroll(-0.95);
initialWindow = FALSE;
}
else if (cgiVarExists("pbt.left2"))
{
relativeScroll(-0.475);
initialWindow = FALSE;
}
else if (cgiVarExists("pbt.left1"))
{
relativeScroll(-0.02);
initialWindow = FALSE;
}
else if (cgiVarExists("pbt.right1"))
{
relativeScroll(0.02);
initialWindow = FALSE;
}
else if (cgiVarExists("pbt.right2"))
{
relativeScroll(0.475);
initialWindow = FALSE;
}
else if (cgiVarExists("pbt.right3"))
{
relativeScroll(0.95);
initialWindow = FALSE;
}
}
dnaUtilOpen();
l=strlen(aa);
/* initialize AA properties */
aaPropertyInit(&hasResFreq);
sfCount = getSuperfamilies2(proteinID);
if (sfCount == 0)
{
sfCount = getSuperfamilies(proteinID);
}
if (mrnaID != NULL)
{
if (kgVersion == KG_III)
{
doExonTrack = FALSE;
sqlSafefFrag(cond_str, sizeof(cond_str), "spId='%s'", proteinID);
kgId = sqlGetField(database, "kgXref", "kgId", cond_str);
if (kgId != NULL)
{
sqlSafefFrag(cond_str, sizeof(cond_str), "name='%s'", kgId);
kgPep = sqlGetField(database, "knownGenePep", "seq", cond_str);
//printf("<pre><br>%s", kgPep);fflush(stdout);
if (kgPep != NULL)
{
if (strstr(protDbName, "proteins") != NULL)
{
protDbDate = strstr(protDbName, "proteins") + strlen("proteins");
safef(uniProtDbName, sizeof(uniProtDbName),"sp%s", protDbDate);
sqlSafefFrag(cond_str, sizeof(cond_str), "acc='%s'", proteinID);
protPep = sqlGetField(uniProtDbName, "protein", "val", cond_str);
//printf("<br>%s\n", protPep);fflush(stdout);
if (protPep != NULL)
{
if (sameWord(kgPep, protPep))
{
//printf("<br>MATCH!\n");fflush(stdout);
sqlSafefFrag(cond_str, sizeof(cond_str), "qName='%s'", kgId);
answer = sqlGetField(database, kgProtMapTableName,
"qName", cond_str);
if (answer != NULL)
{
/* NOTE: passing in kgId instead of proteinID because
kgProtMap2's qName uses kgId instead of
protein display ID */
getExonInfo(kgId, &exCount, &chrom, &strand);
assert(exCount > 0);
doExonTrack = TRUE;
}
}
/*
else
{
chp1 = kgPep;
printf("<br>");
chp2 = protPep;
ll = strlen(kgPep);
if (strlen(protPep) < ll) ll= strlen(protPep);
for (i=0; i<ll; i++)
{
if (*chp1 != *chp2)
{
printf("%c", *chp1);
}
else
{
printf(".");
}
chp1++; chp2++;
}
}
//printf("</pre>");fflush(stdout);
*/
}
}
}
}
}
else
{
doExonTrack = TRUE;
getExonInfo(proteinID, &exCount, &chrom, &strand);
assert(exCount > 0);
}
/* do the following only if pbTracks called doTracks() */
if (initialWindow && IAmPbTracks)
{
prevGBOffsetSav = calPrevGB(exCount, chrom, strand, l, yOffp, proteinID, mrnaID);
trackOrigOffset = prevGBOffsetSav;
if (trackOrigOffset > (protSeqLen*pbScale - 600))
trackOrigOffset = protSeqLen*pbScale - 600;
/* prevent negative value */
if (trackOrigOffset < 0) trackOrigOffset = 0;
}
/* if this if for PDF/Postscript, the trackOrigOffset is already calculated previously,
use the saved value */
if (psOutput != NULL)
{
trackOrigOffset = atoi(cartOptionalString(cart, "pbt.trackOffset"));
}
}
/*printf("<br>%d %d<br>%d %d\n", prevGBStartPos, prevGBEndPos,
blockGenomeStartPositive[exCount-1], blockGenomeStartPositive[0]); fflush(stdout);
*/
if (strand == '-')
{
if ((prevGBStartPos <= blockGenomeStartPositive[exCount-1]) && (prevGBEndPos >= blockGenomeStartPositive[0]))
{
showPrevGBPos = FALSE;
}
}
else
{
if ((prevGBStartPos <= blockGenomeStartPositive[0]) && (prevGBEndPos >= blockGenomeStartPositive[exCount-1]))
{
showPrevGBPos = FALSE;
}
}
if ((cgiOptionalString("aaOrigOffset") != NULL) && scaleButtonPushed)
{
trackOrigOffset = atoi(cgiOptionalString("aaOrigOffset"))*pbScale;
}
pixWidth = 160+ protSeqLen*pbScale;
if (pixWidth > MAX_PB_PIXWIDTH)
{
pixWidth = MAX_PB_PIXWIDTH;
}
if ((protSeqLen*pbScale - trackOrigOffset) < MAX_PB_PIXWIDTH)
{
pixWidth = protSeqLen*pbScale - trackOrigOffset + 160;
}
if (pixWidth < 550) pixWidth = 550;
insideWidth = pixWidth-gfxBorder;
if (proteinInSupportedGenome)
{
pixHeight = 250;
}
else
{
pixHeight = 215;
}
if (sfCount > 0) pixHeight = pixHeight + 20;
/* make room for individual residues display */
if (pbScale >=6) pixHeight = pixHeight + 20;
if (pbScale >=18) pixHeight = pixHeight + 30;
if (psOutput)
{
vg = vgOpenPostScript(pixWidth, pixHeight, psOutput);
suppressHtml = TRUE;
hideControls = TRUE;
}
else
{
trashDirFile(&gifTn, "pbt", "pbt", ".png");
vg = vgOpenPng(pixWidth, pixHeight, gifTn.forCgi, FALSE);
}
/* Put up horizontal scroll controls. */
hWrites("Move ");
hButton("pbt.left3", "<<<");
hButton("pbt.left2", " <<");
hButton("pbt.left1", " < ");
hButton("pbt.right1", " > ");
hButton("pbt.right2", ">> ");
hButton("pbt.right3", ">>>");
hPrintf("     ");
/* Put up scaling controls. */
hPrintf("Current scale: ");
if (pbScale == 1) hPrintf("1/6 ");
if (pbScale == 3) hPrintf("1/2 ");
if (pbScale == 6) hPrintf("FULL ");
if (pbScale == 22) hPrintf("DNA ");
hPrintf("    Rescale to ");
hPrintf("<INPUT TYPE=SUBMIT NAME=\"pbScale\" VALUE=\"1/6\">\n");
hPrintf("<INPUT TYPE=SUBMIT NAME=\"pbScale\" VALUE=\"1/2\">\n");
hPrintf("<INPUT TYPE=SUBMIT NAME=\"pbScale\" VALUE=\"FULL\">\n");
if (kgVersion == KG_III)
{
/* for KG III, the protein has to exist in the kgProtMap2 table
(which will turn on doExonTrack flag)
to provide the genomic position data needed for DNA sequence display */
if ((proteinInSupportedGenome) && (doExonTrack))
hPrintf("<INPUT TYPE=SUBMIT NAME=\"pbScale\" VALUE=\"DNA\">\n");
}
else
{
if (proteinInSupportedGenome)
hPrintf("<INPUT TYPE=SUBMIT NAME=\"pbScale\" VALUE=\"DNA\">\n");
}
hPrintf("<FONT SIZE=1><BR><BR></FONT>\n");
g_vg = vg;
pbRed = vgFindColorIx(g_vg, 0xf9, 0x51, 0x59);
pbBlue = vgFindColorIx(g_vg, 0x00, 0x00, 0xd0);
bkgColor = vgFindColorIx(vg, 255, 254, 232);
vgBox(vg, 0, 0, insideWidth, pixHeight, bkgColor);
/* Start up client side map. */
hPrintf("<MAP Name=%s>\n", mapName);
vgSetClip(vg, 0, gfxBorder, insideWidth, pixHeight - 2*gfxBorder);
/* start drawing indivisual tracks */
doAAScale(l, yOffp, 1);
if (pbScale >= 6) doResidues(aa, l, yOffp);
if (pbScale >= 18) doDnaTrack(chrom, strand, exCount, l, yOffp);
if ((mrnaID != NULL) && showPrevGBPos)
{
doPrevGB(exCount, chrom, strand, l, yOffp, proteinID, mrnaID);
}
if (mrnaID != NULL)
{
if (doExonTrack) doExon(exCount, chrom, l, yOffp, proteinID, mrnaID);
}
doCharge(aa, l, yOffp);
doHydrophobicity(aa, l, yOffp);
doCysteines(aa, l, yOffp);
if (sfCount > 0) doSuperfamily(ensPepName, sfCount, yOffp);
if (hasResFreq) doAnomalies(aa, l, yOffp);
doAAScale(l, yOffp, -1);
vgClose(&vg);
/* Finish map and save out picture and tell html file about it. */
hPrintf("</MAP>\n");
/* put tracks image here */
hPrintf(
"\n<IMG SRC=\"%s\" BORDER=1 WIDTH=%d HEIGHT=%d USEMAP=#%s><BR>",
gifTn.forCgi, pixWidth, pixHeight, mapName);
if (proteinInSupportedGenome)
{
hPrintf("<A HREF=\"../goldenPath/help/pbTracksHelpFiles/pbTracksHelp.shtml#tracks\" TARGET=_blank>");
}
else
{
if (hIsGsidServer())
{
hPrintf("<A HREF=\"../goldenPath/help/pbTracksHelpFiles/pbGsid/pbTracksHelp.shtml#tracks\" TARGET=_blank>");
}
else
{
hPrintf("<A HREF=\"../goldenPath/help/pbTracksHelpFiles/pbTracksHelp.shtml#tracks\" TARGET=_blank>");
}
}
hPrintf("Explanation of Protein Tracks</A><br>");
safef(trackOffset, sizeof(trackOffset), "%d", trackOrigOffset);
cgiMakeHiddenVar("trackOffset", trackOffset);
/* remember where the AA base origin is so that it can be passed to next PB page */
aaOrigOffset = trackOrigOffset/pbScale;
safef(aaOrigOffsetStr, sizeof(aaOrigOffsetStr), "%d", aaOrigOffset);
cgiMakeHiddenVar("aaOrigOffset", aaOrigOffsetStr);
/* save the following state variables, to be used by PDF/Postcript processing */
cartSetString(cart,"pbt.pbScaleStr", pbScaleStr);
cartSetString(cart,"pbt.trackOffset", trackOffset);
cartSaveSession(cart);
fflush(stdout);
}
// TODO move this to a generic re-usable location
void startBackgroundWork(char *exec, char **pWorkUrl)
/* deal with forking off child for background work
* and setting up the trash file for communicating
* from the child to the browser */
{
char *workUrl = NULL;
char hgsid[64];
struct tempName tn;
safef(hgsid, sizeof(hgsid), "%s", cartSessionId(cart));
trashDirFile(&tn, "backGround", hgsid, ".tmp");
workUrl = cloneString(tn.forCgi);
fflush(stdout);
fflush(stderr);
// seems that we need to use the double-fork trick
// to create enough separation between the non-waiting parent
// and the grand-child process. otherwise the OS and Apache are waiting on the child.
int pid = fork();
if (pid == -1)
{
errAbort("can't fork, error %d", errno);
}
if (pid == 0) // child
{
int pid2 = fork();
if (pid2 == -1)
{
errAbort("can't fork, error %d", errno);
}
if (pid2 == 0) // grand child
{
// we need to close or redup to open stdout, stderr, stdin
// in order for apache to break ties with it.
// Will the grandchild cgi still be able to function?
// redirect stdout of child to the trash file for easier use of
// library functions that output html to stdout.
int out = mustOpenFd(tn.forCgi, O_WRONLY | O_CREAT);
fflush(stdout);
dup2(out,STDOUT_FILENO); /* closes STDOUT before setting it back to saved descriptor */
close(out);
// Unfortunately we must create our own stderr log file
char errName[1024];
safef(errName, sizeof errName, "%s.err", tn.forCgi);
int err = mustOpenFd(errName, O_CREAT | O_WRONLY | O_APPEND);
dup2(err, STDERR_FILENO);
close(err);
// stdin input is just empty
int in = mustOpenFd("/dev/null", O_RDONLY);
dup2(in, STDIN_FILENO);
close(in);
// execute so that we will be able to use database and other operations normally.
char execPath[4096];
safef(execPath, sizeof execPath, "%s hgsid=%s", exec, hgsid);
char *args[10];
int numArgs = chopString(execPath, " ", args, 10);
args[numArgs] = NULL;
// by creating a minimal environment and not inheriting from the parent,
// it cause cgiSpoof to run, picking up command-line params as cgi vars.
char *newenviron[] = { "HGDB_CONF=hg.conf", NULL };
int sleepSeconds = 1; // was 5
sleep(sleepSeconds); // Give the foreground process time to write the cart.
execve(args[0], args+1, newenviron);
// SHOULD NOT GET HERE UNLESS EXEC FAILED.
verbose(1,"execve failed for %s\n", exec);
_exit(0); // exit without the usual cleanup which messes up parent's db connections etc.
}
else // child
{
_exit(0); // exit without the usual cleanup which messes up parent's db connections etc.
}
}
else // parent
{
*pWorkUrl = workUrl;
// wait for the exiting child (not grandchild)
int w, status;
do {
w = waitpid(pid, &status, WUNTRACED | WCONTINUED);
if (w == -1)
{
perror("waitpid");
exit(EXIT_FAILURE);
}
if (WIFEXITED(status))
{
if (WEXITSTATUS(status) != 0)
verbose(1, "exited, status=%d\n", WEXITSTATUS(status));
}
else if (WIFSIGNALED(status))
{
verbose(1, "killed by signal %d\n", WTERMSIG(status));
}
else if (WIFSTOPPED(status))
{
verbose(1, "stopped by signal %d\n", WSTOPSIG(status));
}
else if (WIFCONTINUED(status))
{
verbose(1, "continued\n");
}
} while (!WIFEXITED(status) && !WIFSIGNALED(status));
// done waiting for child.
}
}
void blatSeq(char *userSeq, char *organism)
/* Blat sequence user pasted in. */
{
FILE *f;
struct dnaSeq *seqList = NULL, *seq;
struct tempName pslTn, faTn;
int maxSingleSize, maxTotalSize, maxSeqCount;
int minSingleSize = minMatchShown;
char *genome, *db;
char *type = cgiString("type");
char *seqLetters = cloneString(userSeq);
struct serverTable *serve;
int conn;
int oneSize, totalSize = 0, seqCount = 0;
boolean isTx = FALSE;
boolean isTxTx = FALSE;
boolean txTxBoth = FALSE;
struct gfOutput *gvo;
boolean qIsProt = FALSE;
enum gfType qType, tType;
struct hash *tFileCache = gfFileCacheNew();
boolean feelingLucky = cgiBoolean("Lucky");
getDbAndGenome(cart, &db, &genome, oldVars);
if(!feelingLucky)
cartWebStart(cart, db, "%s BLAT Results", trackHubSkipHubName(organism));
/* Load user sequence and figure out if it is DNA or protein. */
if (sameWord(type, "DNA"))
{
seqList = faSeqListFromMemText(seqLetters, TRUE);
uToT(seqList);
isTx = FALSE;
}
else if (sameWord(type, "translated RNA") || sameWord(type, "translated DNA"))
{
seqList = faSeqListFromMemText(seqLetters, TRUE);
uToT(seqList);
isTx = TRUE;
isTxTx = TRUE;
txTxBoth = sameWord(type, "translated DNA");
}
else if (sameWord(type, "protein"))
{
seqList = faSeqListFromMemText(seqLetters, FALSE);
isTx = TRUE;
qIsProt = TRUE;
}
else
{
seqList = faSeqListFromMemTextRaw(seqLetters);
isTx = !seqIsDna(seqList);
if (!isTx)
{
for (seq = seqList; seq != NULL; seq = seq->next)
{
seq->size = dnaFilteredSize(seq->dna);
dnaFilter(seq->dna, seq->dna);
toLowerN(seq->dna, seq->size);
subChar(seq->dna, 'u', 't');
}
}
else
{
for (seq = seqList; seq != NULL; seq = seq->next)
{
seq->size = aaFilteredSize(seq->dna);
aaFilter(seq->dna, seq->dna);
toUpperN(seq->dna, seq->size);
}
qIsProt = TRUE;
}
}
if (seqList != NULL && seqList->name[0] == 0)
{
freeMem(seqList->name);
seqList->name = cloneString("YourSeq");
}
trimUniq(seqList);
/* If feeling lucky only do the first on. */
if(feelingLucky && seqList != NULL)
{
seqList->next = NULL;
}
/* Figure out size allowed. */
maxSingleSize = (isTx ? 10000 : 75000);
maxTotalSize = maxSingleSize * 2.5;
#ifdef LOWELAB
maxSeqCount = 200;
#else
maxSeqCount = 25;
#endif
/* Create temporary file to store sequence. */
trashDirFile(&faTn, "hgSs", "hgSs", ".fa");
faWriteAll(faTn.forCgi, seqList);
/* Create a temporary .psl file with the alignments against genome. */
trashDirFile(&pslTn, "hgSs", "hgSs", ".pslx");
f = mustOpen(pslTn.forCgi, "w");
gvo = gfOutputPsl(0, qIsProt, FALSE, f, FALSE, TRUE);
serve = findServer(db, isTx);
/* Write header for extended (possibly protein) psl file. */
if (isTx)
{
if (isTxTx)
{
qType = gftDnaX;
tType = gftDnaX;
}
else
{
qType = gftProt;
tType = gftDnaX;
}
}
else
{
qType = gftDna;
tType = gftDna;
}
pslxWriteHead(f, qType, tType);
if (qType == gftProt)
{
minSingleSize = 14;
}
else if (qType == gftDnaX)
{
minSingleSize = 36;
}
/* Loop through each sequence. */
for (seq = seqList; seq != NULL; seq = seq->next)
{
printf(" "); fflush(stdout); /* prevent apache cgi timeout by outputting something */
oneSize = realSeqSize(seq, !isTx);
if ((seqCount&1) == 0) // Call bot delay every 2nd time starting with first time
hgBotDelay();
if (++seqCount > maxSeqCount)
{
warn("More than 25 input sequences, stopping at %s.",
seq->name);
break;
}
if (oneSize > maxSingleSize)
{
warn("Sequence %s is %d letters long (max is %d), skipping",
seq->name, oneSize, maxSingleSize);
continue;
}
if (oneSize < minSingleSize)
{
warn("Warning: Sequence %s is only %d letters long (%d is the recommended minimum)",
seq->name, oneSize, minSingleSize);
// we could use "continue;" here to actually enforce skipping,
// but let's give the short sequence a chance, it might work.
// minimum possible length = tileSize+stepSize, so mpl=16 for dna stepSize=5, mpl=10 for protein.
if (qIsProt && oneSize < 1) // protein does not tolerate oneSize==0
continue;
}
totalSize += oneSize;
if (totalSize > maxTotalSize)
{
warn("Sequence %s would take us over the %d letter limit, stopping here.",
seq->name, maxTotalSize);
break;
}
conn = gfConnect(serve->host, serve->port);
if (isTx)
{
gvo->reportTargetStrand = TRUE;
if (isTxTx)
{
gfAlignTransTrans(&conn, serve->nibDir, seq, FALSE, 5,
tFileCache, gvo, !txTxBoth);
if (txTxBoth)
{
reverseComplement(seq->dna, seq->size);
conn = gfConnect(serve->host, serve->port);
gfAlignTransTrans(&conn, serve->nibDir, seq, TRUE, 5,
tFileCache, gvo, FALSE);
}
}
else
{
gfAlignTrans(&conn, serve->nibDir, seq, 5, tFileCache, gvo);
}
}
else
{
gfAlignStrand(&conn, serve->nibDir, seq, FALSE, minMatchShown, tFileCache, gvo);
reverseComplement(seq->dna, seq->size);
conn = gfConnect(serve->host, serve->port);
gfAlignStrand(&conn, serve->nibDir, seq, TRUE, minMatchShown, tFileCache, gvo);
}
gfOutputQuery(gvo, f);
}
carefulClose(&f);
showAliPlaces(pslTn.forCgi, faTn.forCgi, serve->db, qType, tType,
organism, feelingLucky);
if(!feelingLucky)
cartWebEnd();
gfFileCacheFree(&tFileCache);
}
void docIdTidy(char *database, char *docIdDir)
/* docIdTidy - tidy up the docId library by compressing and md5suming where appropriate. */
{
char query[10 * 1024];
struct sqlResult *sr;
char **row;
struct sqlConnection *conn = sqlConnect(database);
struct tempName tn;
trashDirFile(&tn, "docId", "meta", ".txt");
char *tempFile = tn.forCgi;
struct toDoList *toDoList = NULL;
safef(query, sizeof query, "select * from %s", docIdTable);
sr = sqlGetResult(conn, query);
while ((row = sqlNextRow(sr)) != NULL)
{
struct docIdSub *docIdSub = docIdSubLoad(row);
cgiDecode(docIdSub->metaData, docIdSub->metaData, strlen(docIdSub->metaData));
FILE *f = mustOpen(tempFile, "w");
fwrite(docIdSub->metaData, strlen(docIdSub->metaData), 1, f);
fclose(f);
boolean validated;
//printf("metadata is %s\n", docIdSub->metaData);
struct mdbObj *mdbObj = mdbObjsLoadFromFormattedFile(tempFile, &validated);
unlink(tempFile);
char *docIdType = mdbObjFindValue(mdbObj, "type");
char buffer[10 * 1024];
safef(buffer, sizeof buffer, "%d", docIdSub->ix);
char *path = docIdGetPath(buffer, docIdDir, docIdType, NULL);
//docIdDecorate(docIdSub->ix));
printf("path %s\n", path);
struct toDoList *toDoItem = NULL;
if (!fileIsCompressed(path))
{
if (toDoItem == NULL) AllocVar(toDoItem);
printf("foo\n");
toDoItem->needs |= NEEDS_COMPRESSION;
toDoItem->path = path;
toDoItem->docId = docIdSub->ix;
}
printf("docId %d md5sum %s valReport %s\n",docIdSub->ix, docIdSub->md5sum, docIdSub->valReport);
//if (docIdSub->md5sum == NULL)
if (sameString(docIdSub->md5sum, ""))
{
printf("mdsum\n");
if (toDoItem == NULL) AllocVar(toDoItem);
toDoItem->needs |= NEEDS_MD5SUM;
toDoItem->path = path;
toDoItem->docId = docIdSub->ix;
}
//if (docIdSub->valReport == NULL)
if (sameString(docIdSub->valReport,""))
{
printf("report\n");
if (toDoItem == NULL) AllocVar(toDoItem);
toDoItem->needs |= NEEDS_REPORT;
toDoItem->path = path;
toDoItem->docId = docIdSub->ix;
}
if (toDoItem)
slAddHead(&toDoList, toDoItem);
}
sqlFreeResult(&sr);
doCompression(toDoList);
doReports(toDoList);
doMd5Summing(toDoList);
}
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);
}
void makeActiveImagePB(char *psOutput, char *psOutput2)
/* Make image and image map. */
{
char *mapName = "map";
int pixWidth, pixHeight;
struct sqlConnection *conn;
char query[256];
struct sqlResult *sr;
char **row;
int iypos;
char *spDisplayId;
char *oldDisplayId;
conn = sqlConnect(UNIPROT_DB_NAME);
printf("<BR>");
hPrintf("<BR><font size=4><B>Protein: ");
hPrintf("%s</B>", proteinID);
/* Please note the hiv database name is hard wired here.*/
safef(query, sizeof(query),
"select subjId from hivVax003Vax004.gsIdXref where aaSeqId = '%s'", proteinID);
sr = sqlMustGetResult(conn, query);
row = sqlNextRow(sr);
if (row != NULL)
{
printf("<BR>");
hPrintf("<font size=4><B>Subject: ");
hPrintf("<A HREF=\"../cgi-bin/gsidSubj?hgs_subj=%s&submit=Go!\">", row[0]);
hPrintf("%s</A></B><BR>", row[0]);
}
sqlFreeResult(&sr);
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("</font><br>");
protSeq = getAA(proteinID);
if (protSeq == NULL)
{
errAbort("%s is not a current valid entry in UniProt(SWISS-PROT/TrEMBL)\n", proteinID);
}
protSeqLen = strlen(protSeq);
fflush(stdout);
iypos = 15;
doTracks(proteinID, mrnaID, protSeq, &iypos, psOutput);
if (!hTableExists(database, "pbStamp")) goto histDone;
pbScale = 3;
pixWidth = 520;
insideWidth = pixWidth-gfxBorder;
pixHeight = 350;
if (psOutput2)
{
vg2 = vgOpenPostScript(pixWidth, pixHeight, psOutput2);
}
else
{
trashDirFile(&gifTn2, "pbt", "pbt", ".gif");
vg2 = vgOpenGif(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/pbGsid/pbTracksHelp.shtml#histograms\" TARGET=_blank>");
}
hPrintf("Explanation of Protein Property Histograms</A><BR>");
hPrintf("<P>");
histDone:
hPrintf("<P>");
fflush(stdout);
printFASTA(proteinID, protSeq);
}
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();
}
static void generateImgFiles(struct hgdpGeo *geo, char finalEpsFile[PATH_LEN],
char finalPdfFile[PATH_LEN], char finalPngFile[PATH_LEN])
/* Using the frequencies given in geo and the population latitude and longitude
* given above, plot allele frequency pie charts for each population on a world map.
* Work in a temporary trash dir and then move result files to the given paths. */
{
// The Generic Mapping Tools commands must have a writeable ./ and $HOME.
// Use trashDirFile as a directory name, cd to that and work there.
char cwd[PATH_LEN];
if (getcwd(cwd, sizeof(cwd)) == NULL)
errAbort("PATH_LEN (%d) is too short.", PATH_LEN);
struct tempName dirTn;
trashDirFile(&dirTn, "hgc", "hgdpGeo", "");
makeDirs(dirTn.forCgi);
mustChdir(dirTn.forCgi);
char *realHome = getenv("HOME");
setenv("HOME", ".", TRUE);
// Make trash files with coordinate specs for pie charts and full circles:
char rootName[FILENAME_LEN];
splitPath(dirTn.forCgi, NULL, rootName, NULL);
char pieFile[FILENAME_LEN], circleFile[FILENAME_LEN];
safef(pieFile, sizeof(pieFile), "%s_pie.txt", rootName);
safef(circleFile, sizeof(circleFile), "%s_circle.txt", rootName);
FILE *fPie = mustOpen(pieFile, "w");
FILE *fCir = mustOpen(circleFile, "w");
int i;
for (i = 0; i < HGDPGEO_POP_COUNT; i++)
{
FILE *f = (geo->popFreqs[i] == 1.0) ? fCir : fPie;
fprintf(f, "%.4f %.4f 0.65 0 %.4f\n",
pops[i].latitude, pops[i].longitude, geo->popFreqs[i]*360);
}
fclose(fPie);
fclose(fCir);
// Build up an EPS image: copy a baseline EPS (map background), add SNP and allele labels,
// and use ps2xy commands to allele frequency pie charts.
char epsFile[FILENAME_LEN];
safef(epsFile, sizeof(epsFile), "%s.eps", rootName);
char mapBgEps[PATH_LEN];
safef(mapBgEps, sizeof(mapBgEps), "%s/hgcData/hgdpGeoMap.eps", cwd);
copyFile(mapBgEps, epsFile);
FILE *fEps = mustOpen(epsFile, "a");
fprintf(fEps, hgdpGeoLabelFormat, geo->name, geo->name,
geo->ancestralAllele, geo->ancestralAllele, geo->derivedAllele, geo->derivedAllele);
fclose(fEps);
//- run psxy on circle/pie spec trash files for AfrEurAsia and America
char *psxy = cfgOption("hgc.psxyPath");
if (isEmpty(psxy))
errAbort("How did this get called? hg.conf doesn't have hgc.psxyPath.");
runCommandAppend(psxyOrangeAeaCmd, psxy, pieFile, epsFile);
runCommandAppend(psxyPieAeaCmd, psxy, pieFile, epsFile);
runCommandAppend(psxyBlueAeaCmd, psxy, circleFile, epsFile);
runCommandAppend(psxyOrangeAmCmd, psxy, pieFile, epsFile);
runCommandAppend(psxyPieAmCmd, psxy, pieFile, epsFile);
runCommandAppend(psxyBlueAmCmd, psxy, circleFile, epsFile);
// Make PDF and PNG:
struct pipeline *pl;
char pdfFile[FILENAME_LEN], pngFile[FILENAME_LEN];
safef(pdfFile, sizeof(pdfFile), "%s.pdf", rootName);
safef(pngFile, sizeof(pngFile), "%s.png", rootName);
char *ps2pdfCmd[] = {"ps2pdf", epsFile, pdfFile, NULL};
char **cmdsPdf[] = {ps2pdfCmd, NULL};
pl = pipelineOpen(cmdsPdf, pipelineWrite, "/dev/null", NULL);
pipelineClose(&pl);
char *ps2raster = cfgOption("hgc.ps2rasterPath");
char *ghostscript = cfgOption("hgc.ghostscriptPath");
char gsOpt[PATH_LEN];
safef(gsOpt, sizeof(gsOpt), "-G%s", ghostscript);
char *ps2RasterPngCmd[] = {ps2raster, gsOpt, "-P", "-A", "-Tg", "-E150", epsFile, NULL};
char **cmdsPng[] = {ps2RasterPngCmd, NULL};
pl = pipelineOpen(cmdsPng, pipelineRead, "/dev/null", NULL);
pipelineClose(&pl);
// Back to our usual working directory and $HOME:
if (realHome == NULL)
unsetenv("HOME");
else
setenv("HOME", realHome, TRUE);
mustChdir(cwd);
// Move the result files into place:
char tmpPath[PATH_LEN];
safef(tmpPath, sizeof(tmpPath), "%s/%s", dirTn.forCgi, epsFile);
mustRename(tmpPath, finalEpsFile);
safef(tmpPath, sizeof(tmpPath), "%s/%s", dirTn.forCgi, pdfFile);
mustRename(tmpPath, finalPdfFile);
safef(tmpPath, sizeof(tmpPath), "%s/%s", dirTn.forCgi, pngFile);
mustRename(tmpPath, finalPngFile);
// Clean up the temporary working directory (trash cleaner script doesn't
// remove empty dirs:
removeDir(dirTn.forCgi);
}
