struct slName *getTablesForField(struct sqlConnection *conn,
char *splitPrefix, char *table, char *splitSuffix)
/* Get tables that match field. */
{
struct slName *list = NULL, *el;
if (splitPrefix != NULL || splitSuffix != NULL)
{
char query[256], **row;
struct sqlResult *sr;
safef(query, sizeof(query), "show tables like '%s%s%s'",
emptyForNull(splitPrefix), table, emptyForNull(splitSuffix));
sr = sqlGetResult(conn, query);
while ((row = sqlNextRow(sr)) != NULL)
{
el = slNameNew(row[0]);
slAddHead(&list, el);
}
sqlFreeResult(&sr);
slReverse(&list);
}
if (list == NULL)
{
if (sqlTableExists(conn, table))
list = slNameNew(table);
}
return list;
}
struct composite *makeCompositeList(struct encode2Manifest *manList, struct hash *metaHash)
/* Return a list of composites with everything on manList */
{
struct composite *comp, *compList = NULL;
struct hash *compHash = hashNew(0);
char compName[256];
struct encode2Manifest *man;
for (man = manList; man != NULL; man = man->next)
{
char *realComp = tagVal(man, metaHash, "composite");
if (realComp != NULL)
safef(compName, sizeof(compName), "%s", realComp);
else
{
char *lab = emptyForNull(tagVal(man, metaHash, "lab"));
char *dataType = emptyForNull(tagVal(man, metaHash, "dataType"));
safef(compName, sizeof(compName), "comp%s%s", lab, dataType);
}
comp = hashFindVal(compHash, compName);
if (comp == NULL)
{
AllocVar(comp);
comp->name = cloneString(compName);
hashAdd(compHash, compName, comp);
slAddTail(&compList, comp);
}
struct slRef *manRef = slRefNew(man);
slAddTail(&comp->manRefList, manRef);
}
hashFree(&compHash);
return compList;
}
void writeSeriesList(char *fileName, struct series *list)
/* Write out list to file in tab separated format, adding initial id field. */
{
FILE *f = mustOpen(fileName, "w");
int id = 0;
struct series *series;
for (series = list; series != NULL; series = series->next)
{
fprintf(f, "%d\t%s\t%s\t%s\n", ++id, series->name,
emptyForNull(series->dataType), emptyForNull(series->grantee));
}
carefulClose(&f);
}
void printLeafTrackList(FILE *f, char *indent, struct view *view, struct composite *comp,
struct slName *varList, struct taggedFile *tfList, char *type)
/* Print list of low level tracks under view */
{
struct taggedFile *tf;
for (tf = tfList; tf != NULL; tf = tf->next)
{
if (sameString(tf->manifest->outputType, view->name))
{
fprintf(f, "%strack t%d\n", indent, ++trackId);
fprintf(f, "%sparent %s\n", indent, view->trackName);
fprintf(f, "%stype %s\n", indent, type);
fprintf(f, "%ssubGroups view=%s", indent, view->name);
struct slName *var;
for (var = varList; var != NULL; var = var->next)
{
char *val = metaTagValFindVal(tf->tagList, var->name);
if (val != NULL)
fprintf(f, " %s=%s", var->name, val);
}
fprintf(f, "\n");
fprintf(f, "%sshortLabel", indent);
for (var = varList; var != NULL; var = var->next)
{
char *val = metaTagValFindVal(tf->tagList, var->name);
if (val != NULL)
fprintf(f, " %s", val);
}
fprintf(f, "\n");
char *lab = emptyForNull(metaTagValFindVal(tf->tagList, "lab"));
char *dataType = emptyForNull(metaTagValFindVal(tf->tagList, "dataType"));
fprintf(f, "%slongLabel %s %s", indent, lab, dataType);
boolean gotOne = FALSE;
for (var = varList; var != NULL; var = var->next)
{
char *val = metaTagValFindVal(tf->tagList, var->name);
if (val != NULL)
{
if (gotOne)
fprintf(f, ",");
else
gotOne = TRUE;
fprintf(f, " %s %s", var->name, val);
}
}
fprintf(f, "\n");
fprintf(f, "%sbigDataUrl %s\n", indent, tf->manifest->fileName);
fprintf(f, "\n");
}
}
}
void edwMakePlateFileNameAndPath(int edwFileId, char *submitFileName,
char licensePlate[edwMaxPlateSize], char edwFile[PATH_LEN], char serverPath[PATH_LEN])
/* Convert file id to local file name, and full file path. Make any directories needed
* along serverPath. */
{
/* Preserve suffix. Give ourselves up to two suffixes. */
int nameSize = strlen(submitFileName);
char *suffix = lastMatchCharExcept(submitFileName, submitFileName + nameSize, '.', '/');
if (suffix != NULL)
{
char *secondSuffix = lastMatchCharExcept(submitFileName, suffix, '.', '/');
if (secondSuffix != NULL)
suffix = secondSuffix;
}
suffix = emptyForNull(suffix);
/* Figure out edw file name, starting with license plate. */
edwMakeLicensePlate(edwLicensePlatePrefix, edwFileId, licensePlate, edwMaxPlateSize);
/* Figure out directory and make any components not already there. */
char edwDir[PATH_LEN];
edwDirForTime(edwNow(), edwDir);
char uploadDir[PATH_LEN];
safef(uploadDir, sizeof(uploadDir), "%s%s", edwRootDir, edwDir);
makeDirsOnPath(uploadDir);
/* Figure out full file names */
safef(edwFile, PATH_LEN, "%s%s%s", edwDir, licensePlate, suffix);
safef(serverPath, PATH_LEN, "%s%s", edwRootDir, edwFile);
}
struct dyString *readAndReplaceTableName(char *fileName, char *table)
/* Read file into string. While doing so strip any leading comments
* and insist that the first non-comment line contain the words
* "create table" followed by a table name. Replace the table name,
* and copy the rest of the file verbatem. */
{
struct lineFile *lf = lineFileOpen(fileName, TRUE);
struct dyString *dy = dyStringNew(0);
char *line, *word;
if (!lineFileNextReal(lf, &line))
errAbort("No real lines in %s\n", fileName);
word = nextWord(&line);
if (!sameWord(word, "create"))
errAbort("Expecting first word in file to be CREATE. Got %s", word);
word = nextWord(&line);
if (word == NULL || !sameWord(word, "table"))
errAbort("Expecting second word in file to be table. Got %s", emptyForNull(word));
word = nextWord(&line);
if (word == NULL)
errAbort("Expecting table name on same line as CREATE TABLE");
sqlDyStringPrintf(dy, "CREATE TABLE %s ", table);
if (line != NULL)
dyStringAppend(dy, line);
dyStringAppendC(dy, '\n');
while (lineFileNext(lf, &line, NULL))
{
dyStringAppend(dy, line);
dyStringAppendC(dy, '\n');
}
lineFileClose(&lf);
return dy;
}
void transformJobTable(struct sqlConnection *conn, struct edwAnalysisJob *jobList, char *outName)
/* Transform edwAnalysisJob to edwAnalysis */
{
FILE *f = mustOpen(outName, "w");
struct edwAnalysisJob *job;
for (job = jobList; job != NULL; job = job->next)
{
struct eapJob j = {0};
j.id = job->id;
j.commandLine = sqlEscapeTabFileString(job->commandLine);
assert(job->commandLine);
j.startTime = job->startTime;
j.endTime = job->endTime;
j.stderr = sqlEscapeTabFileString(emptyForNull(job->stderr));
j.returnCode = job->returnCode;
j.cpusRequested = max(1, job->cpusRequested);
char buf[16];
if (isEmpty(job->parasolId))
{
safef(buf, sizeof(buf), "%d", job->pid);
j.parasolId = buf;
}
else
j.parasolId = job->parasolId;
eapJobTabOut(&j, f);
freez(&j.commandLine);
freez(&j.stderr);
}
carefulClose(&f);
}
static void showTableDataRows(struct fieldedTable *table, int pageSize, int maxLenField,
struct hash *tagOutputWrappers, void *wrapperContext)
/* Render data rows into HTML */
{
int count = 0;
struct fieldedRow *row;
boolean isNum[table->fieldCount];
int i;
for (i=0; i<table->fieldCount; ++i)
isNum[i] = fieldedTableColumnIsNumeric(table, i);
for (row = table->rowList; row != NULL; row = row->next)
{
if (++count > pageSize)
break;
printf("<TR>\n");
int fieldIx = 0;
for (fieldIx=0; fieldIx<table->fieldCount; ++fieldIx)
{
char shortVal[maxLenField+1];
char *longVal = emptyForNull(row->row[fieldIx]);
char *val = longVal;
int valLen = strlen(val);
if (maxLenField > 0 && maxLenField < valLen)
{
if (valLen > maxLenField)
{
memcpy(shortVal, val, maxLenField-3);
shortVal[maxLenField-3] = 0;
strcat(shortVal, "...");
val = shortVal;
}
}
if (isNum[fieldIx])
webPrintLinkCellRightStart();
else
webPrintLinkCellStart();
boolean printed = FALSE;
if (tagOutputWrappers != NULL && !isEmpty(val))
{
char *field = table->fields[fieldIx];
webTableOutputWrapperType *printer = hashFindVal(tagOutputWrappers, field);
if (printer != NULL)
{
printer(table, row, field, longVal, val, wrapperContext);
printed = TRUE;
}
}
if (!printed)
printf("%s", val);
webPrintLinkCellEnd();
}
printf("</TR>\n");
}
}
struct raField *raFieldNew(char *name, char *val, struct lm *lm)
/* Return new raField. */
{
struct raField *field;
lmAllocVar(lm, field);
field->name = lmCloneString(lm, name);
val = emptyForNull(skipLeadingSpaces(val));
field->val = lmCloneString(lm, val);
return field;
}
static struct hash *getCoveredTables(struct joiner *joiner, char *db,
struct sqlConnection *conn)
/* Get list of tables covered in database. */
{
struct hash *hash = hashNew(0);
struct joinerIgnore *ig;
struct slName *spec;
struct joinerSet *js;
struct joinerField *jf;
/* First put in all the ignored tables. */
for (ig = joiner->tablesIgnored; ig != NULL; ig = ig->next)
{
if (slNameInList(ig->dbList, db))
{
for (spec = ig->tableList; spec != NULL; spec = spec->next)
{
verbose(3,"ignoreTable: '%s'\n", spec->name);
addTablesLike(hash, conn, spec->name);
}
}
}
/* Now put in tables that are in one of the identifiers. */
for (js = joiner->jsList; js != NULL; js = js->next)
{
for (jf = js->fieldList; jf != NULL; jf = jf->next)
{
if (slNameInList(jf->dbList, db))
{
char spec[512];
safef(spec, sizeof(spec), "%s%s%s",
emptyForNull(jf->splitPrefix), jf->table,
emptyForNull(jf->splitSuffix));
addTablesLike(hash, conn, spec);
verbose(4,"ident: '%s', table: '%s'\n", js->name, spec);
}
}
}
return hash;
}
static void writeInfo(FILE *infoFh, struct gffGroup *group)
/* write a row for a GTF group from the info file */
{
// scan lineList for group and protein ids
struct gffLine *ll;
char *geneId = NULL, *proteinId = NULL, *geneName = NULL, *transcriptName = NULL;
for (ll = group->lineList; ll != NULL; ll = ll->next)
{
saveName(&geneId, ll->geneId);
saveName(&proteinId, ll->proteinId);
saveName(&geneName, ll->geneName);
saveName(&transcriptName, ll->transcriptName);
}
fprintf(infoFh, "%s\t%s\t%s\t%s\t%d\t%d\t%c\t%s\t%s\t%s\n",
group->name, emptyForNull(geneId), group->source,
group->seq, group->start, group->end, group->strand,
emptyForNull(proteinId), emptyForNull(geneName),
emptyForNull(transcriptName));
}
static struct rqlEval rqlEvalArrayIx(struct rqlParse *p, void *record, RqlEvalLookup lookup,
struct lm *lm)
/* Handle parse tree generated by an indexed array. */
{
struct rqlParse *array = p->children;
struct rqlParse *index = array->next;
struct rqlEval arrayVal = rqlLocalEval(array, record, lookup, lm);
struct rqlEval indexVal = rqlLocalEval(index, record, lookup, lm);
struct rqlEval res;
res.type = rqlTypeString;
res.val.s = emptyForNull(lmCloneSomeWord(lm, arrayVal.val.s, indexVal.val.i));
return res;
}
void output(int depth, struct rqlStatement *rql, struct tagStorm *tags, struct tagStanza *stanza)
/* Output stanza according to clOut */
{
char *format = clOut;
if (sameString(format, "ra"))
{
if (stanza->children == NULL)
{
struct slName *field;
for (field = rql->fieldList; field != NULL; field = field->next)
{
char *val = tagFindVal(stanza, field->name);
if (val != NULL)
printf("%s\t%s\n", field->name, val);
}
printf("\n");
}
}
else if (sameString(format, "tab"))
{
if (stanza->children == NULL)
{
struct slName *field;
char *connector = "";
for (field = rql->fieldList; field != NULL; field = field->next)
{
char *val = emptyForNull(tagFindVal(stanza, field->name));
printf("%s%s", connector, val);
connector = "\t";
}
printf("\n");
}
}
else if (sameString(format, "tags"))
{
struct slName *field;
for (field = rql->fieldList; field != NULL; field = field->next)
{
char *val = tagFindLocalVal(stanza, field->name);
if (val != NULL)
{
repeatCharOut(stdout, '\t', depth);
printf("%s\t%s\n", field->name, val);
}
}
printf("\n");
}
else
errAbort("Unrecognized format %s", format);
}
void getUrl(struct sqlConnection *conn)
/* Put up URL. */
{
edwMustGetUserFromEmail(conn, userEmail);
printf("Please enter a URL for a validated manifest file:<BR>");
printf("URL ");
cgiMakeTextVar("url", emptyForNull(cgiOptionalString("url")), 80);
cgiMakeButton("submitUrl", "submit");
printf("<BR>\n");
cgiMakeCheckBox("update", FALSE);
printf(" Update information associated with files that have already been uploaded.");
printf("<BR>Submission by %s", userEmail);
edwPrintLogOutButton();
}
static void refLinkUpdate(struct sqlConnection *conn, struct gbStatus* status)
/* Update the refLink table for the current entry */
{
int geneId;
char *gen = emptyForNull(raFieldCurVal("gen"));
char *pro = emptyForNull(raFieldCurVal("pro"));
gen = sqlEscapeString2(alloca(2*strlen(gen)+1), gen);
pro = sqlEscapeString2(alloca(2*strlen(pro)+1), pro);
/* can either have locus id (old locus link db) or gene id, or both,
* in which case the geneId is used */
geneId = (raGeneId != 0) ? raGeneId : raLocusLinkId;
if (status->stateChg & GB_NEW)
sqlUpdaterAddRow(refLinkUpd, "%s\t%s\t%s\t%s\t%u\t%u\t%u\t%u",
gen, pro, raAcc, raProtAcc, raFieldCurId("gen"),
raFieldCurId("pro"), geneId, raOmimId);
else if (status->stateChg & GB_META_CHG)
sqlUpdaterModRow(refLinkUpd, 1, "name='%s', product='%s', protAcc='%s', "
"geneName=%u, prodName=%u, locusLinkId=%u, "
"omimId=%u where mrnaAcc='%s'",
gen, pro, raProtAcc, raFieldCurId("gen"),
raFieldCurId("pro"), geneId, raOmimId, raAcc);
}
struct customTrack *chromGraphParser(char *genomeDb, struct customPp *cpp,
char *formatType, char *markerType, char *columnLabels,
char *name, char *description, struct hash *settings,
boolean report)
/* Parse out a chromGraph file (not including any track lines) */
{
char *minVal = hashFindVal(settings, "minVal");
char *maxVal = hashFindVal(settings, "maxVal");
/* Get first lines of track. If track empty then
* might as well return NULL here. */
struct slName *preview = getPreview(cpp, 10);
if (preview == NULL)
return NULL;
/* Figure out format type - scanning preview if it isn't well defined. */
struct sqlConnection *conn = hAllocConn(genomeDb);
int colCount;
if (sameString(formatType, cgfFormatGuess))
{
if (!figureOutFormat(preview, &formatType, &colCount))
errAbort("Can't figure out format for chromGraph track %s",
emptyForNull(name));
}
hashMayRemove(settings, "formatType");
/* Now that we know format can count columns and determine how to
* chop up lines. */
colCount = countColumns(preview, formatType);
Chopper chopper = getChopper(formatType);
/* Figure out marker type - scanning marker column of preview if it isn't
* well defined. */
if (sameString(markerType, cgfMarkerGuess))
{
markerType = guessMarkerType(preview, chopper, conn, colCount);
if (markerType == NULL)
errAbort("Can't figure out marker column type for chromGraph track %s",
emptyForNull(name));
}
hashMayRemove(settings, "markerType");
/* Figure out if columns are labeled in file, using preview if needed. */
if (sameString(columnLabels, cgfColLabelGuess))
{
if (firstRowConsistentWithData(preview->name, chopper, colCount))
columnLabels = cgfColLabelNumbered;
else
columnLabels = cgfColLabelFirstRow;
}
hashMayRemove(settings, "columnLabels");
returnPreview(cpp, &preview);
boolean labelsInData = sameString(columnLabels, cgfColLabelFirstRow);
/* Load data into list of labeled temp files. */
struct labeledFile *fileEl, *fileList;
fileList = parseToLabeledFiles(cpp, colCount, formatType, markerType,
labelsInData, conn, report);
saveLabeledFileList(fileList);
/* Create a customTrack for each element in file list. */
struct customTrack *outputTracks = NULL;
for (fileEl = fileList; fileEl != NULL; fileEl = fileEl->next)
{
struct customTrack *track;
AllocVar(track);
struct trackDb *tdb = customTrackTdbDefault();
track->tdb = tdb;
/* Figure out short and long names and type*/
char shortLabel[128];
char longLabel[512];
if (name == NULL)
name = track->tdb->shortLabel;
if (description == NULL)
description = track->tdb->longLabel;
if (colCount > 1 || labelsInData)
{
safef(shortLabel, sizeof(shortLabel), "%s %s", name, fileEl->label);
safef(longLabel, sizeof(longLabel), "%s %s", description, fileEl->label);
}
else
{
safef(shortLabel, sizeof(shortLabel), "%s", name);
safef(longLabel, sizeof(longLabel), "%s", description);
}
tdb->shortLabel = cloneString(shortLabel);
tdb->longLabel = cloneString(longLabel);
tdb->type = "chromGraph";
tdb->track = customTrackTableFromLabel(tdb->shortLabel);
tdb->table = cloneString(tdb->track);
track->dbTableName = NULL;
/* Create settings */
struct dyString *dy = dyStringNew(0);
dyStringAppend(dy, hashToRaString(settings));
dyStringPrintf(dy, "binaryFile %s\n", fileEl->fileName);
dyStringPrintf(dy, "type %s\n", tdb->type);
dyStringPrintf(dy, "tdbType %s\n", tdb->type); /* Needed if outside factory */
if (minVal == NULL)
dyStringPrintf(dy, "minVal %g\n", fileEl->minVal);
if (maxVal == NULL)
dyStringPrintf(dy, "maxVal %g\n", fileEl->maxVal);
tdb->settings = dyStringCannibalize(&dy);
/* Add to list. */
slAddHead(&outputTracks, track);
}
hFreeConn(&conn);
slReverse(&outputTracks);
return outputTracks;
}
void writeMdbListAsResults(struct mdbObj *mdbList, char *fileName)
/* Write selected fields in list in tab-separated result format to file. */
{
FILE *f = mustOpen(fileName, "w");
struct mdbObj *mdb;
int id = 0;
for (mdb = mdbList; mdb != NULL; mdb = mdb->next)
{
char *experiment = NULL;
char *replicate = NULL;
char *objType = NULL;
char *fileName = NULL;
char *md5sum = NULL;
char *tableName = NULL;
char *view = NULL;
char *dateSubmitted = NULL;
char *dateResubmitted = NULL;
char *dateUnrestricted = NULL;
struct mdbVar *v;
for (v = mdb->vars; v != NULL; v = v->next)
{
char *var = v->var, *val = v->val;
if (sameString(var, "expId"))
experiment = val;
else if (sameString(var, "replicate"))
replicate = val;
else if (sameString(var, "objType"))
objType = val;
else if (sameString(var, "fileName"))
{
fileName = val;
/* Do surgery on file name, cutting off second comma separated index file. */
char *comma = strchr(fileName, ',');
if (comma != NULL)
*comma = 0;
}
else if (sameString(var, "md5sum"))
{
md5sum = val;
char *comma = strchr(fileName, ',');
if (comma != NULL)
*comma = 0;
}
else if (sameString(var, "tableName"))
tableName = val;
else if (sameString(var, "view"))
view = val;
else if (sameString(var, "dateSubmitted"))
dateSubmitted = val;
else if (sameString(var, "dateResubmitted"))
dateResubmitted = val;
else if (sameString(var, "dateUnrestricted"))
dateUnrestricted = val;
}
if (objType != NULL && !sameString(objType, "composite"))
{
if (experiment != NULL)
{
fprintf(f, "%d\t", ++id);
fprintf(f, "%s\t", emptyForNull(experiment));
fprintf(f, "%s\t", emptyForNull(replicate));
fprintf(f, "%s\t", emptyForNull(view));
fprintf(f, "%s\t", emptyForNull(objType));
fprintf(f, "%s\t", emptyForNull(fileName));
fprintf(f, "%s\t", emptyForNull(md5sum));
fprintf(f, "%s\t", emptyForNull(tableName));
fprintf(f, "%s\t", emptyForNull(dateSubmitted));
fprintf(f, "%s\t", emptyForNull(dateResubmitted));
fprintf(f, "%s\n", emptyForNull(dateUnrestricted));
}
else
{
if (!startsWith("wgEncodeUmassWengTfbsValid", mdb->obj)) // These validation files not associated with encode experiment
warn("No experiment for %s", mdb->obj);
}
}
}
carefulClose(&f);
}
static void synonymPrint(struct section *section,
struct sqlConnection *conn, char *id)
/* Print out SwissProt comments - looking up typeId/commentVal. */
{
char *protAcc = getSwissProtAcc(conn, spConn, id);
char *spDisplayId;
char *refSeqAcc = "";
char *mrnaAcc = "";
char *oldDisplayId;
char condStr[255];
char *kgProteinID;
char *parAcc; /* parent accession of a variant splice protein */
char *chp;
if (isRgdGene(conn))
{
rgdGene2SynonymPrint(section,conn, id);
return;
}
if (sqlTablesExist(conn, "kgAlias"))
printAlias(id, conn);
if (sameWord(genome, "Zebrafish"))
{
char *xrefTable = "ensXRefZfish";
char *geneIdCol = "ensGeneId";
/* get Gene Symbol and RefSeq accession from Zebrafish-specific */
/* cross-reference table */
printGeneSymbol(id, xrefTable, geneIdCol, conn);
refSeqAcc = getRefSeqAcc(id, xrefTable, geneIdCol, conn);
hPrintf("<B>ENSEMBL ID:</B> %s", id);
}
else
{
char query[256];
char *toRefTable = genomeOptionalSetting("knownToRef");
if (toRefTable != NULL && sqlTableExists(conn, toRefTable))
{
safef(query, sizeof(query), "select value from %s where name='%s'", toRefTable,
id);
refSeqAcc = emptyForNull(sqlQuickString(conn, query));
}
if (sqlTableExists(conn, "kgXref"))
{
safef(query, sizeof(query), "select mRNA from kgXref where kgID='%s'", id);
mrnaAcc = emptyForNull(sqlQuickString(conn, query));
}
if (sameWord(genome, "C. elegans"))
hPrintf("<B>WormBase ID:</B> %s<BR>", id);
else
hPrintf("<B>UCSC ID:</B> %s<BR>", id);
}
if (refSeqAcc[0] != 0)
{
hPrintf("<B>RefSeq Accession: </B> <A HREF=\"");
printOurRefseqUrl(stdout, refSeqAcc);
hPrintf("\">%s</A><BR>\n", refSeqAcc);
}
else if (mrnaAcc[0] != 0)
{
safef(condStr, sizeof(condStr), "acc = '%s'", mrnaAcc);
if (sqlGetField(database, "gbCdnaInfo", "acc", condStr) != NULL)
{
hPrintf("<B>Representative RNA: </B> <A HREF=\"");
printOurMrnaUrl(stdout, mrnaAcc);
hPrintf("\">%s</A><BR>\n", mrnaAcc);
}
else
/* do not show URL link if it is not found in gbCdnaInfo */
{
hPrintf("<B>Representative RNA: %s </B>", mrnaAcc);
}
}
if (protAcc != NULL)
{
kgProteinID = cloneString("");
if (hTableExists(sqlGetDatabase(conn), "knownGene")
&& (isNotEmpty(cartOptionalString(cart, hggChrom)) &&
differentWord(cartOptionalString(cart, hggChrom),"none")))
{
safef(condStr, sizeof(condStr), "name = '%s' and chrom = '%s' and txStart=%s and txEnd=%s",
id, cartOptionalString(cart, hggChrom),
cartOptionalString(cart, hggStart),
cartOptionalString(cart, hggEnd));
kgProteinID = sqlGetField(database, "knownGene", "proteinID", condStr);
}
hPrintf("<B>Protein: ");
if (strstr(kgProteinID, "-") != NULL)
{
parAcc = cloneString(kgProteinID);
chp = strstr(parAcc, "-");
*chp = '\0';
/* show variant splice protein and the UniProt link here */
hPrintf("<A HREF=\"http://www.uniprot.org/uniprot%s\" "
"TARGET=_blank>%s</A></B>, splice isoform of ",
kgProteinID, kgProteinID);
hPrintf("<A HREF=\"http://www.uniprot.org/uniprot/%s\" "
"TARGET=_blank>%s</A></B>\n",
parAcc, parAcc);
}
else
{
hPrintf("<A HREF=\"http://www.uniprot.org/uniprot/%s\" "
"TARGET=_blank>%s</A></B>\n",
protAcc, protAcc);
}
/* show SWISS-PROT display ID if it is different than the accession ID */
/* but, if display name is like: Q03399 | Q03399_HUMAN, then don't show display name */
spDisplayId = spAnyAccToId(spConn, protAcc);
if (spDisplayId == NULL)
{
errAbort("<br>%s seems to no longer be a valid protein ID in our latest UniProtKB DB.", protAcc);
}
if (strstr(spDisplayId, protAcc) == 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)
&& !sameWord(protAcc, oldDisplayId))
{
hPrintf(" or %s", oldDisplayId);
}
}
hPrintf(")<BR>\n");
}
}
printCcds(id, conn);
}
static void bigBedClick(char *fileName, struct trackDb *tdb,
char *item, int start, int end, int bedSize)
/* Handle click in generic bigBed track. */
{
boolean showUrl = FALSE;
char *chrom = cartString(cart, "c");
/* Open BigWig file and get interval list. */
struct bbiFile *bbi = bigBedFileOpen(fileName);
struct lm *lm = lmInit(0);
int ivStart = start, ivEnd = end;
if (start == end)
{
// item is an insertion; expand the search range from 0 bases to 2 so we catch it:
ivStart = max(0, start-1);
ivEnd++;
}
struct bigBedInterval *bbList = bigBedIntervalQuery(bbi, chrom, ivStart, ivEnd, 0, lm);
/* Get bedSize if it's not already defined. */
if (bedSize == 0)
{
bedSize = bbi->definedFieldCount;
showUrl = TRUE;
}
char *scoreFilter = cartOrTdbString(cart, tdb, "scoreFilter", NULL);
int minScore = 0;
if (scoreFilter)
minScore = atoi(scoreFilter);
/* Find particular item in list - matching start, and item if possible. */
boolean found = FALSE;
boolean firstTime = TRUE;
struct bigBedInterval *bb;
for (bb = bbList; bb != NULL; bb = bb->next)
{
if (!(bb->start == start && bb->end == end))
continue;
if (bedSize > 3)
{
char *name = cloneFirstWordByTab(bb->rest);
boolean match = sameString(name, item);
freez(&name);
if (!match)
continue;
}
found = TRUE;
if (firstTime)
printf("<BR>\n");
int seq1Seq2Fields = 0;
// check for seq1 and seq2 in columns 7+8 (eg, pairedTagAlign)
boolean seq1Seq2 = sameOk(trackDbSetting(tdb, BASE_COLOR_USE_SEQUENCE), "seq1Seq2");
if (seq1Seq2 && bedSize == 6)
seq1Seq2Fields = 2;
char *fields[bedSize+seq1Seq2Fields];
char startBuf[16], endBuf[16];
char *rest = cloneString(bb->rest);
int bbFieldCount = bigBedIntervalToRow(bb, chrom, startBuf, endBuf, fields,
bedSize+seq1Seq2Fields);
if (bbFieldCount != bedSize+seq1Seq2Fields)
{
errAbort("Disagreement between trackDb field count (%d) and %s fieldCount (%d)",
bedSize, fileName, bbFieldCount);
}
struct bed *bed = bedLoadN(fields, bedSize);
if (bedSize >= 6 && scoreFilter && bed->score < minScore)
continue;
if (showUrl && (bedSize >= 4))
printCustomUrl(tdb, item, TRUE);
bedPrintPos(bed, bedSize, tdb);
// display seq1 and seq2
if (seq1Seq2 && bedSize+seq1Seq2Fields == 8)
printf("<table><tr><th>Sequence 1</th><th>Sequence 2</th></tr>"
"<tr><td> %s </td><td> %s </td></tr></table>", fields[6], fields[7]);
else if (isNotEmpty(rest))
{
char *restFields[256];
int restCount = chopTabs(rest, restFields);
int restBedFields = bedSize - 3;
if (restCount > restBedFields)
{
if (0 == extraFieldsPrint(tdb,NULL,restFields + restBedFields,restCount - restBedFields))
{
int i;
char label[20];
safef(label, sizeof(label), "nonBedFieldsLabel");
printf("<B>%s </B>",
trackDbSettingOrDefault(tdb, label, "Non-BED fields:"));
for (i = restBedFields; i < restCount; i++)
printf("%s%s", (i > 0 ? "\t" : ""), restFields[i]);
printf("<BR>\n");
}
}
}
if (isCustomTrack(tdb->track))
{
time_t timep = bbiUpdateTime(bbi);
printBbiUpdateTime(&timep);
}
}
if (!found)
{
printf("No item %s starting at %d\n", emptyForNull(item), start);
}
lmCleanup(&lm);
bbiFileClose(&bbi);
}
void processRefSeq(char *database, char *faFile, char *raFile, char *pslFile, char *loc2refFile,
char *pepFile, char *mim2locFile)
/* hgRefSeqMrna - Load refSeq mRNA alignments and other info into
* refSeqGene table. */
{
struct lineFile *lf;
struct hash *raHash, *rsiHash = newHash(0);
struct hash *loc2mimHash = newHash(0);
struct refSeqInfo *rsiList = NULL, *rsi;
char *s, *line, *row[5];
int wordCount, dotMod = 0;
int noLocCount = 0;
int rsiCount = 0;
int noProtCount = 0;
struct psl *psl;
struct sqlConnection *conn = hgStartUpdate(database);
struct hash *productHash = loadNameTable(conn, "productName", 16);
struct hash *geneHash = loadNameTable(conn, "geneName", 16);
char *kgName = "refGene";
FILE *kgTab = hgCreateTabFile(".", kgName);
FILE *productTab = hgCreateTabFile(".", "productName");
FILE *geneTab = hgCreateTabFile(".", "geneName");
FILE *refLinkTab = hgCreateTabFile(".", "refLink");
FILE *refPepTab = hgCreateTabFile(".", "refPep");
FILE *refMrnaTab = hgCreateTabFile(".", "refMrna");
struct exon *exonList = NULL, *exon;
char *answer;
char cond_str[200];
/* Make refLink and other tables table if they don't exist already. */
sqlMaybeMakeTable(conn, "refLink", refLinkTableDef);
sqlUpdate(conn, "NOSQLINJ delete from refLink");
sqlMaybeMakeTable(conn, "refGene", refGeneTableDef);
sqlUpdate(conn, "NOSQLINJ delete from refGene");
sqlMaybeMakeTable(conn, "refPep", refPepTableDef);
sqlUpdate(conn, "NOSQLINJ delete from refPep");
sqlMaybeMakeTable(conn, "refMrna", refMrnaTableDef);
sqlUpdate(conn, "NOSQLINJ delete from refMrna");
/* Scan through locus link to omim ID file and put in hash. */
{
char *row[2];
printf("Scanning %s\n", mim2locFile);
lf = lineFileOpen(mim2locFile, TRUE);
while (lineFileRow(lf, row))
{
hashAdd(loc2mimHash, row[1], intToPt(atoi(row[0])));
}
lineFileClose(&lf);
}
/* Scan through .ra file and make up start of refSeqInfo
* objects in hash and list. */
printf("Scanning %s\n", raFile);
lf = lineFileOpen(raFile, TRUE);
while ((raHash = hashNextRa(lf)) != NULL)
{
if (clDots > 0 && ++dotMod == clDots )
{
dotMod = 0;
dotOut();
}
AllocVar(rsi);
slAddHead(&rsiList, rsi);
if ((s = hashFindVal(raHash, "acc")) == NULL)
errAbort("No acc near line %d of %s", lf->lineIx, lf->fileName);
rsi->mrnaAcc = cloneString(s);
if ((s = hashFindVal(raHash, "siz")) == NULL)
errAbort("No siz near line %d of %s", lf->lineIx, lf->fileName);
rsi->size = atoi(s);
if ((s = hashFindVal(raHash, "gen")) != NULL)
rsi->geneName = cloneString(s);
//!!!else
//!!! warn("No gene name for %s", rsi->mrnaAcc);
if ((s = hashFindVal(raHash, "cds")) != NULL)
parseCds(s, 0, rsi->size, &rsi->cdsStart, &rsi->cdsEnd);
else
rsi->cdsEnd = rsi->size;
if ((s = hashFindVal(raHash, "ngi")) != NULL)
rsi->ngi = atoi(s);
rsi->geneNameId = putInNameTable(geneHash, geneTab, rsi->geneName);
s = hashFindVal(raHash, "pro");
if (s != NULL)
rsi->productName = cloneString(s);
rsi->productNameId = putInNameTable(productHash, productTab, s);
hashAdd(rsiHash, rsi->mrnaAcc, rsi);
freeHashAndVals(&raHash);
}
lineFileClose(&lf);
if (clDots) printf("\n");
/* Scan through loc2ref filling in some gaps in rsi. */
printf("Scanning %s\n", loc2refFile);
lf = lineFileOpen(loc2refFile, TRUE);
while (lineFileNext(lf, &line, NULL))
{
char *mrnaAcc;
if (line[0] == '#')
continue;
wordCount = chopTabs(line, row);
if (wordCount < 5)
errAbort("Expecting at least 5 tab-separated words line %d of %s",
lf->lineIx, lf->fileName);
mrnaAcc = row[1];
mrnaAcc = accWithoutSuffix(mrnaAcc);
if (mrnaAcc[2] != '_')
warn("%s is and odd name %d of %s",
mrnaAcc, lf->lineIx, lf->fileName);
if ((rsi = hashFindVal(rsiHash, mrnaAcc)) != NULL)
{
rsi->locusLinkId = lineFileNeedNum(lf, row, 0);
rsi->omimId = ptToInt(hashFindVal(loc2mimHash, row[0]));
rsi->proteinAcc = cloneString(accWithoutSuffix(row[4]));
}
}
lineFileClose(&lf);
/* Report how many seem to be missing from loc2ref file.
* Write out knownInfo file. */
printf("Writing %s\n", "refLink.tab");
for (rsi = rsiList; rsi != NULL; rsi = rsi->next)
{
++rsiCount;
if (rsi->locusLinkId == 0)
++noLocCount;
if (rsi->proteinAcc == NULL)
++noProtCount;
fprintf(refLinkTab, "%s\t%s\t%s\t%s\t%u\t%u\t%u\t%u\n",
emptyForNull(rsi->geneName),
emptyForNull(rsi->productName),
emptyForNull(rsi->mrnaAcc),
emptyForNull(rsi->proteinAcc),
rsi->geneNameId, rsi->productNameId,
rsi->locusLinkId, rsi->omimId);
}
if (noLocCount)
printf("Missing locusLinkIds for %d of %d\n", noLocCount, rsiCount);
if (noProtCount)
printf("Missing protein accessions for %d of %d\n", noProtCount, rsiCount);
/* Process alignments and write them out as genes. */
lf = pslFileOpen(pslFile);
dotMod = 0;
while ((psl = pslNext(lf)) != NULL)
{
if (hashFindVal(rsiHash, psl->qName) != NULL)
{
if (clDots > 0 && ++dotMod == clDots )
{
dotMod = 0;
dotOut();
}
sqlSafefFrag(cond_str, sizeof cond_str, "extAC='%s'", psl->qName);
answer = sqlGetField(proteinDB, "spXref2", "displayID", cond_str);
if (answer == NULL)
{
fprintf(stderr, "%s NOT FOUND.\n", psl->qName);
fflush(stderr);
}
if (answer != NULL)
{
struct genePred *gp = NULL;
exonList = pslToExonList(psl);
fprintf(kgTab, "%s\t%s\t%c\t%d\t%d\t",
psl->qName, psl->tName, psl->strand[0], psl->tStart, psl->tEnd);
rsi = hashMustFindVal(rsiHash, psl->qName);
gp = genePredFromPsl(psl, rsi->cdsStart, rsi->cdsEnd, genePredStdInsertMergeSize);
if (!gp)
errAbort("Cannot convert psl (%s) to genePred.\n", psl->qName);
fprintf(kgTab, "%d\t%d\t", gp->cdsStart, gp->cdsEnd);
fprintf(kgTab, "%d\t", slCount(exonList));
fflush(kgTab);
for (exon = exonList; exon != NULL; exon = exon->next)
fprintf(kgTab, "%d,", exon->start);
fprintf(kgTab, "\t");
for (exon = exonList; exon != NULL; exon = exon->next)
fprintf(kgTab, "%d,", exon->end);
fprintf(kgTab, "\n");
slFreeList(&exonList);
}
}
else
{
fprintf(stderr, "%s found in psl, but not in .fa or .ra data files.\n", psl->qName);
fflush(stderr);
}
}
if (clDots) printf("\n");
if (!clTest)
{
writeSeqTable(pepFile, refPepTab, FALSE, TRUE);
writeSeqTable(faFile, refMrnaTab, FALSE, FALSE);
}
carefulClose(&kgTab);
carefulClose(&productTab);
carefulClose(&geneTab);
carefulClose(&refLinkTab);
carefulClose(&refPepTab);
carefulClose(&refMrnaTab);
if (!clTest)
{
printf("Loading database with %s\n", kgName);
fflush(stdout);
hgLoadTabFile(conn, ".", kgName, NULL);
printf("Loading database with %s\n", "productName");
fflush(stdout);
hgLoadTabFile(conn, ".", "productName", NULL);
printf("Loading database with %s\n", "geneName");
fflush(stdout);
hgLoadTabFile(conn, ".", "geneName", NULL);
printf("Loading database with %s\n", "refLink");
fflush(stdout);
hgLoadTabFile(conn, ".", "refLink", NULL);
printf("Loading database with %s\n", "refPep");
fflush(stdout);
hgLoadTabFile(conn, ".", "refPep", NULL);
printf("Loading database with %s\n", "refMrna");
fflush(stdout);
hgLoadTabFile(conn, ".", "refMrna", NULL);
}
}
void txGeneFromBed(char *inBed, char *inPicks, char *ucscFa, char *uniProtFa, char *refPepFa, char *outKg)
/* txGeneFromBed - Convert from bed to knownGenes format table (genePred + uniProt ID). */
{
/* Load protein sequence into hashes */
struct hash *uniProtHash = faReadAllIntoHash(uniProtFa, dnaUpper);
struct hash *ucscProtHash = faReadAllIntoHash(ucscFa, dnaUpper);
struct hash *refProtHash =faReadAllIntoHash(refPepFa, dnaUpper);
/* Load picks into hash. We don't use cdsPicksLoadAll because empty fields
* cause that autoSql-generated routine problems. */
struct hash *pickHash = newHash(18);
struct cdsPick *pick;
struct lineFile *lf = lineFileOpen(inPicks, TRUE);
char *row[CDSPICK_NUM_COLS];
while (lineFileRowTab(lf, row))
{
pick = cdsPickLoad(row);
hashAdd(pickHash, pick->name, pick);
}
/* Load in bed */
struct bed *bed, *bedList = bedLoadNAll(inBed, 12);
/* Do reformatting and write output. */
FILE *f = mustOpen(outKg, "w");
for (bed = bedList; bed != NULL; bed = bed->next)
{
char *protAcc = NULL;
if (bed->thickStart < bed->thickEnd)
{
pick = hashMustFindVal(pickHash, bed->name);
struct dnaSeq *spSeq = NULL, *uniSeq = NULL, *refPep = NULL, *ucscSeq;
ucscSeq = hashMustFindVal(ucscProtHash, bed->name);
if (pick->swissProt[0])
spSeq = hashMustFindVal(uniProtHash, pick->swissProt);
if (pick->uniProt[0])
uniSeq = hashMustFindVal(uniProtHash, pick->uniProt);
if (pick->refProt[0])
refPep = hashMustFindVal(refProtHash, pick->refProt);
/* First we look for an exact match between the ucsc protein and
* something from swissProt/uniProt. */
if (spSeq != NULL && sameString(ucscSeq->dna, spSeq->dna))
protAcc = pick->swissProt;
if (protAcc == NULL && uniSeq != NULL && sameString(ucscSeq->dna, uniSeq->dna))
protAcc = pick->uniProt;
if (protAcc == NULL && refPep != NULL && sameString(ucscSeq->dna, refPep->dna))
{
protAcc = cloneString(pick->refProt);
chopSuffix(protAcc);
}
if (protAcc == NULL)
{
if (pick->uniProt[0])
protAcc = pick->uniProt;
else
{
protAcc = cloneString(pick->refProt);
chopSuffix(protAcc);
}
}
}
outputKg(bed, emptyForNull(protAcc), f);
}
carefulClose(&f);
}
