void txGeneAlias(char *genomeDb, char *uniProtDb, char *xrefFile,
char *evFile, char *oldToNew, char *aliasFile, char *protAliasFile)
/* txGeneAlias - Make kgAlias and kgProtAlias tables.. */
{
/* Read and hash oldToNew */
struct hash *newToOldHash = loadNewToOldHash(oldToNew);
/* Load evidence into hash */
struct hash *evHash = newHash(18);
struct txRnaAccs *ev, *evList = txRnaAccsLoadAll(evFile);
for (ev = evList; ev != NULL; ev = ev->next)
hashAdd(evHash, ev->name, ev);
/* Open connections to our databases */
struct sqlConnection *gConn = sqlConnect(genomeDb);
struct sqlConnection *uConn = sqlConnect(uniProtDb);
struct sqlResult *sr;
char **row;
char query[256];
/* Open files. */
struct lineFile *lf = lineFileOpen(xrefFile, TRUE);
FILE *fAlias = mustOpen(aliasFile, "w");
FILE *fProt = mustOpen(protAliasFile, "w");
/* Stream through xref file, which has much of the info we need,
* and which contains a line for each gene. */
char *words[KGXREF_NUM_COLS];
while (lineFileRowTab(lf, words))
{
/* Load the xref, and output most of it's fields as aliases. */
struct kgXref *x = kgXrefLoad(words);
char *id = x->kgID;
outAlias(fAlias, id, x->kgID);
outAlias(fAlias, id, x->mRNA);
outAlias(fAlias, id, x->spID);
outAlias(fAlias, id, x->spDisplayID);
outAlias(fAlias, id, x->geneSymbol);
outAlias(fAlias, id, x->refseq);
outAlias(fAlias, id, x->protAcc);
char *old = hashFindVal(newToOldHash, id);
if (old != NULL)
outAlias(fAlias, id, old);
/* If we've got a uniProt ID, use that to get more info from uniProt. */
char *acc = x->spID;
if ((acc[0] != 0) && (acc = spLookupPrimaryAccMaybe(uConn, acc)) != NULL)
{
/* Get current accession and output a bunch of easy protein aliases. */
outProt(fProt, id, acc, acc);
outProt(fProt, id, acc, x->spDisplayID);
outProt(fProt, id, acc, x->geneSymbol);
outProt(fProt, id, acc, x->protAcc);
if (old != NULL)
outProt(fProt, id, acc, old);
/* Throw in old swissProt accessions. */
sqlSafef(query, sizeof(query), "select val from otherAcc where acc = '%s'", acc);
sr = sqlGetResult(uConn, query);
while ((row = sqlNextRow(sr)) != NULL)
{
outAlias(fAlias, id, row[0]);
outProt(fProt, id, acc, row[0]);
}
/* Throw in gene names that SwissProt knows about */
struct slName *gene, *geneList = spGenes(uConn, acc);
for (gene = geneList; gene != NULL; gene = gene->next)
{
outAlias(fAlias, id, gene->name);
outProt(fProt, id, acc, gene->name);
}
slFreeList(&geneList);
}
/* Throw in gene names from genbank. */
/* At some point we may want to restrict this to the primary transcript in a cluster. */
ev = hashFindVal(evHash, id);
if (ev != NULL)
{
int i;
for (i=0; i<ev->accCount; ++i)
{
sqlSafef(query, sizeof(query), "select geneName from gbCdnaInfo where acc='%s'", acc);
int nameId = sqlQuickNum(gConn, query);
if (nameId != 0)
{
char name[64];
sqlSafef(query, sizeof(query), "select name from geneName where id=%d", nameId);
if (sqlQuickQuery(gConn, query, name, sizeof(name)))
outAlias(fAlias, id, name);
}
}
}
kgXrefFree(&x);
}
carefulClose(&fAlias);
carefulClose(&fProt);
}
void spTest(char *database, char *someAcc)
/* spTest - Test out sp library.. */
{
struct sqlConnection *conn = sqlConnect(database);
char *acc, *id, *binomial, *common;
struct slName *geneList, *gene, *accList, *n, *list;
struct slName *nameList, *name, *keyList, *key, *typeList, *type;
struct spFeature *featList, *feat;
struct spCitation *citeList, *cite;
char *ret = NULL;
int taxon;
int classId = 0, typeId = 0, refId = 0;
printf("input: %s\n", someAcc);
acc = spLookupPrimaryAcc(conn, someAcc);
printf("primary accession: %s\n", acc);
id = spAccToId(conn, acc);
printf("SwissProt id: %s\n", id);
printf("acc from id: %s\n", spIdToAcc(conn, id));
ret = spOrganelle(conn, acc);
printf("organelle: %s\n", (ret == NULL) ? "(null)" : ret);
printf("isCurated: %d\n", spIsCurated(conn, acc));
printf("aaSize: %d\n", spAaSize(conn,acc));
printf("molWeight: %d\n", spMolWeight(conn,acc));
printf("createDate: %s\n", spCreateDate(conn,acc));
printf("seqDate: %s\n", spSeqDate(conn,acc));
printf("annDate: %s\n", spAnnDate(conn,acc));
printf("description: %s\n", spDescription(conn, acc));
taxon = spTaxon(conn, acc);
printf("taxon: %d\n", taxon);
binomial = spTaxonToBinomial(conn, taxon);
printf("first scientific name: %s\n", binomial);
common = spTaxonToCommon(conn, taxon);
printf("first common name: %s\n", common);
printf("taxon from sci: %d\n", spBinomialToTaxon(conn, binomial));
printf("taxon from common: %d\n", spCommonToTaxon(conn, common));
printf("all scientific names:");
nameList = spBinomialNames(conn, acc);
for (name = nameList; name != NULL; name = name->next)
printf(" %s,", name->name);
printf("\n");
printf("gene(s):");
geneList = spGenes(conn,acc);
for (gene=geneList; gene != NULL; gene = gene->next)
printf(" %s,", gene->name);
printf("\n");
for (gene=geneList; gene != NULL; gene = gene->next)
{
accList = spGeneToAccs(conn, gene->name, 0);
printf(" any %s:", gene->name);
for (n = accList; n != NULL; n = n->next)
printf(" %s,", n->name);
printf("\n");
slFreeList(&accList);
printf(" %s %s:", common, gene->name);
accList = spGeneToAccs(conn, gene->name, taxon);
for (n = accList; n != NULL; n = n->next)
printf(" %s,", n->name);
printf("\n");
slFreeList(&accList);
}
slFreeList(&geneList);
printf("keyword(s):");
keyList = spKeywords(conn, acc);
for (key = keyList; key != NULL; key = key->next)
printf(" %s,", key->name);
printf("\n");
for (key = keyList; key != NULL; key = key->next)
{
accList = spKeywordSearch(conn, key->name, taxon);
printPartialList(common, key->name, accList, 4);
slFreeList(&accList);
break; /* This is a little slow, once is enough. */
}
for (key = keyList; key != NULL; key = key->next)
{
accList = spKeywordSearch(conn, key->name, 0);
printPartialList("all", key->name, accList, 4);
slFreeList(&accList);
break; /* This is a little slow, once is enough. */
}
slFreeList(&keyList);
printf("All comments:\n");
list = slComments(conn, acc, NULL);
for (n = list; n != NULL; n = n->next)
printf(" %s\n", n->name);
slFreeList(&list);
typeList = slCommentTypes(conn);
for (type = typeList; type != NULL; type = type->next)
{
list = slComments(conn, acc, type->name);
if (list != NULL)
{
printf("%s comments:\n", type->name);
for (n = list; n != NULL; n = n->next)
printf(" %s\n", n->name);
slFreeList(&list);
}
}
slFreeList(&typeList);
list = spEmblAccs(conn, acc);
printf("GenBank/EMBL:");
for (n = list; n != NULL; n = n->next)
printf(" %s,", n->name);
printf("\n");
if (list != NULL)
printf("acc from %s: %s\n",
list->name, spAccFromEmbl(conn, list->name));
slFreeList(&list);
list = spPdbAccs(conn, acc);
printf("PDB:");
for (n = list; n != NULL; n = n->next)
printf(" %s,", n->name);
printf("\n");
featList = spFeatures(conn, acc, 0, 0);
printf("All features:\n");
for (feat = featList; feat != NULL; feat = feat->next)
{
printFeat(conn, feat);
classId = feat->featureClass;
typeId = feat->featureType;
}
slFreeList(&featList);
if (classId != 0 && typeId != 0)
{
printf("%s class features:\n", spFeatureClassName(conn, classId));
featList = spFeatures(conn, acc, classId, 0);
for (feat = featList; feat != NULL; feat = feat->next)
printFeat(conn, feat);
slFreeList(&featList);
printf("%s type features:\n", spFeatureTypeName(conn, typeId));
featList = spFeatures(conn, acc, 0, typeId);
for (feat = featList; feat != NULL; feat = feat->next)
printFeat(conn, feat);
slFreeList(&featList);
printf("same class & type features:\n");
featList = spFeatures(conn, acc, classId, typeId);
for (feat = featList; feat != NULL; feat = feat->next)
printFeat(conn, feat);
slFreeList(&featList);
printf("class loop: %d->%s->%d\n", classId,
spFeatureClassName(conn, classId),
spFeatureClassId(conn, spFeatureClassName(conn, classId)));
printf("type loop: %d->%s->%d\n", typeId,
spFeatureTypeName(conn, typeId),
spFeatureTypeId(conn, spFeatureTypeName(conn, typeId)));
}
citeList = spCitations(conn, acc);
for (cite = citeList; cite != NULL; cite = cite->next)
{
refId = cite->reference;
printf("title: %s\n", spRefTitle(conn, refId));
printf("authors:");
list = spRefAuthors(conn, refId);
for (n = list; n != NULL; n = n->next)
printf(" %s, ", n->name);
printf("\n");
slFreeList(&list);
printf("location: %s\n", spRefCite(conn, refId));
printf("pubMed: %s\n", spRefPubMed(conn, refId));
}
if (refId != 0)
{
printf("other accs associated with last reference:\n\t");
list = spRefToAccs(conn, refId);
printPartialList("", "", list, 6);
slFreeList(&list);
}
sqlDisconnect(&conn);
}
void spDbAddVarSplice(char *database, char *inFile, char *outDir)
/* spDbAddVarSplice - This adds information on the varient splices to the sp/uniProt database. */
{
struct sqlConnection *conn = sqlConnect(database);
char query[256];
makeDir(outDir);
FILE *varProtein = openToWrite(outDir, "varProtein.txt");
FILE *varAcc = openToWrite(outDir, "varAcc.txt");
FILE *varDisplayId = openToWrite(outDir, "varDisplayId.txt");
FILE *varAccToTaxon = openToWrite(outDir, "varAccToTaxon.txt");
FILE *varDescription = openToWrite(outDir, "varDescription.txt");
FILE *varGene = openToWrite(outDir, "varGene.txt");
FILE *varGeneLogic = openToWrite(outDir, "varGeneLogic.txt");
struct lineFile *lf = lineFileOpen(inFile, TRUE);
aaSeq seq;
ZeroVar(&seq);
while (faPepSpeedReadNext(lf, &seq.dna, &seq.size, &seq.name))
{
char *row[4];
char *name = seq.name;
if (startsWith("sp|", name)) // Skip over sp| introduced Aug 2009
name += 3;
int rowSize = chopString(name, "-|", row, ArraySize(row));
if (rowSize != 3)
errAbort("Expecting name to be in format accession-N|DISP_ID, got %s\n", name);
char *acc = row[0];
char *version = row[1];
char *displayId = row[2];
int accLen = strlen(acc);
int verLen = strlen(version);
int displayIdLen = strlen(displayId);
/* Do some tests. */
if ((accLen != 6 && accLen != 10) || isdigit(acc[0]) || !isdigit(acc[accLen-1]))
errAbort("wierd accession %s before line %d of %s", acc, lf->lineIx, lf->fileName);
if (!isdigit(version[0]) || verLen > 4)
errAbort("wierd version %s before line %d of %s", version, lf->lineIx, lf->fileName);
if (countChars(displayId, '_') != 1 || displayIdLen < 6 || displayIdLen > 16)
errAbort("wierd displayId %s before line %d of %s", displayId, lf->lineIx, lf->fileName);
if (accLen + 1 + verLen >= sizeof(SpAcc))
errAbort("Need to increase size of SpAcc in spDb.h because of %s-%s - need %d characters but only have %lu", acc, version, accLen + 1 + verLen, sizeof(SpAcc));
/* Print out parsed results. */
fprintf(varAcc, "%s-%s\t%s\t%s\n", acc, version, acc, version);
fprintf(varProtein, "%s-%s\t%s\n", acc, version, seq.dna);
fprintf(varDisplayId, "%s-%s\t%s-%s\n", acc, version, acc, version);
/* Look up taxon of base protein and use it to write to varAccToTaxon table. */
int taxon = spTaxon(conn, acc);
fprintf(varAccToTaxon, "%s-%s\t%d\n", acc, version, taxon);
/*Transfer description. */
char *description = spDescription(conn, acc);
fprintf(varDescription, "%s-%s\t%s\n", acc, version, description);
freez(&description);
/* Transfer gene logic. */
sqlSafef(query, sizeof(query), "select val from geneLogic where acc = '%s'", acc);
char *geneLogic = sqlQuickString(conn, query);
if (geneLogic != NULL)
fprintf(varGeneLogic, "%s-%s\t%s\n", acc, version, geneLogic);
freez(&geneLogic);
/* Transfer genes. */
struct slName *gene, *geneList = spGenes(conn, acc);
for (gene = geneList; gene != NULL; gene = gene->next)
fprintf(varGene, "%s-%s\t%s\n", acc, version, gene->name);
slFreeList(&geneList);
}
carefulClose(&varAcc);
carefulClose(&varProtein);
carefulClose(&varDisplayId);
carefulClose(&varAccToTaxon);
carefulClose(&varDescription);
carefulClose(&varGene);
carefulClose(&varGeneLogic);
sqlDisconnect(&conn);
}
void txGeneXref(char *genomeDb, char *uniProtDb, char *genePredFile, char *infoFile, char *pickFile,
char *evFile, char *outFile)
/* txGeneXref - Make kgXref type table for genes.. */
{
/* Load picks into hash. We don't use cdsPicksLoadAll because empty fields
* cause that autoSql-generated routine problems. */
struct hash *pickHash = newHash(18);
struct hash *geneToProtHash = makeGeneToProtHash(genePredFile);
struct cdsPick *pick;
struct lineFile *lf = lineFileOpen(pickFile, TRUE);
char *row[CDSPICK_NUM_COLS];
while (lineFileRowTab(lf, row))
{
pick = cdsPickLoad(row);
removePickVersions(pick);
hashAdd(pickHash, pick->name, pick);
}
/* Load evidence into hash */
struct hash *evHash = newHash(18);
struct txRnaAccs *ev, *evList = txRnaAccsLoadAll(evFile);
for (ev = evList; ev != NULL; ev = ev->next)
hashAdd(evHash, ev->name, ev);
/* Open connections to our databases */
struct sqlConnection *gConn = sqlConnect(genomeDb);
struct sqlConnection *uConn = sqlConnect(uniProtDb);
/* Read in info file, and loop through it to make out file. */
struct txInfo *info, *infoList = txInfoLoadAll(infoFile);
FILE *f = mustOpen(outFile, "w");
for (info = infoList; info != NULL; info = info->next)
{
char *kgID = info->name;
char *mRNA = "";
char *spID = "";
char *spDisplayID = "";
char *geneSymbol = NULL;
char *refseq = "";
char *protAcc = "";
char *description = NULL;
char query[256];
char *proteinId = hashMustFindVal(geneToProtHash, info->name);
boolean isAb = sameString(info->category, "antibodyParts");
pick = hashFindVal(pickHash, info->name);
ev = hashFindVal(evHash, info->name);
if (pick != NULL)
{
/* Fill in the relatively straightforward fields. */
refseq = pick->refSeq;
if (info->orfSize > 0)
{
protAcc = pick->refProt;
spID = proteinId;
if (sameString(protAcc, spID))
spID = pick->uniProt;
if (spID[0] != 0)
spDisplayID = spAnyAccToId(uConn, spID);
}
/* Fill in gene symbol and description from refseq if possible. */
if (refseq[0] != 0)
{
struct sqlResult *sr;
safef(query, sizeof(query), "select name,product from refLink where mrnaAcc='%s'",
refseq);
sr = sqlGetResult(gConn, query);
char **row = sqlNextRow(sr);
if (row != NULL)
{
geneSymbol = cloneString(row[0]);
if (!sameWord("unknown protein", row[1]))
description = cloneString(row[1]);
}
sqlFreeResult(&sr);
}
/* If need be try uniProt for gene symbol and description. */
if (spID[0] != 0 && (geneSymbol == NULL || description == NULL))
{
char *acc = spLookupPrimaryAcc(uConn, spID);
if (description == NULL)
description = spDescription(uConn, acc);
if (geneSymbol == NULL)
{
struct slName *nameList = spGenes(uConn, acc);
if (nameList != NULL)
geneSymbol = cloneString(nameList->name);
slFreeList(&nameList);
}
}
}
/* If it's an antibody fragment use that as name. */
if (isAb)
{
geneSymbol = cloneString("abParts");
description = cloneString("Parts of antibodies, mostly variable regions.");
isAb = TRUE;
}
if (ev == NULL)
{
mRNA = cloneString("");
if (!isAb)
{
errAbort("%s is %s but not %s\n", info->name, infoFile, evFile);
}
}
else
{
mRNA = cloneString(ev->primary);
chopSuffix(mRNA);
}
/* Still no joy? Try genbank RNA records. */
if (geneSymbol == NULL || description == NULL)
{
if (ev != NULL)
{
int i;
for (i=0; i<ev->accCount; ++i)
{
char *acc = ev->accs[i];
chopSuffix(acc);
if (geneSymbol == NULL)
{
safef(query, sizeof(query),
"select geneName.name from gbCdnaInfo,geneName "
"where geneName.id=gbCdnaInfo.geneName and gbCdnaInfo.acc = '%s'", acc);
geneSymbol = sqlQuickString(gConn, query);
if (geneSymbol != NULL)
{
if (sameString(geneSymbol, "n/a"))
geneSymbol = NULL;
}
}
if (description == NULL)
{
safef(query, sizeof(query),
"select description.name from gbCdnaInfo,description "
"where description.id=gbCdnaInfo.description "
"and gbCdnaInfo.acc = '%s'", acc);
description = sqlQuickString(gConn, query);
if (description != NULL)
{
if (sameString(description, "n/a"))
description = NULL;
}
}
}
}
}
if (geneSymbol == NULL)
geneSymbol = mRNA;
if (description == NULL)
description = mRNA;
/* Get rid of some characters that will cause havoc downstream. */
stripChar(geneSymbol, '\'');
subChar(geneSymbol, '<', '[');
subChar(geneSymbol, '>', ']');
/* Abbreviate geneSymbol if too long */
if (strlen(geneSymbol) > 40)
strcpy(geneSymbol+37, "...");
fprintf(f, "%s\t", kgID);
fprintf(f, "%s\t", mRNA);
fprintf(f, "%s\t", spID);
fprintf(f, "%s\t", spDisplayID);
fprintf(f, "%s\t", geneSymbol);
fprintf(f, "%s\t", refseq);
fprintf(f, "%s\t", protAcc);
fprintf(f, "%s\n", description);
}
carefulClose(&f);
}