struct segment *parseSegment(char *fileName, int start, int end, char *retSeqName)
/* Read in a genscan file into segment. */
{
struct lineFile *lf = lineFileOpen(fileName, TRUE);
struct segment *seg;
char *line;
struct genScanFeature *gsfList = NULL, *gsf;
struct genScanGene *gsg;
char *words[2];
if (!lineFileNext(lf, &line, NULL))
errAbort("%s is empty", fileName);
if (!startsWith("GENSCAN ", line))
errAbort("%s is not a GENSCAN output file", fileName);
if (retSeqName != NULL)
{
line = mustSkipTo(lf, "Sequence");
if (chopLine(line, words) < 2)
errAbort("Expecting sequence name line %d of %s", lf->lineIx, lf->fileName);
strcpy(retSeqName, words[1]);
}
mustSkipTo(lf, "Predicted genes/exons");
mustSkipTo(lf, "Gn.Ex");
mustSkipTo(lf, "-----");
AllocVar(seg);
seg->start = start;
seg->end = end;
for (;;)
{
if (!lineFileNext(lf, &line, NULL))
break;
line = skipLeadingSpaces(line);
if (line == NULL || line[0] == 0)
continue;
if (!isdigit(line[0]))
{
lineFileReuse(lf);
break;
}
gsf = parseGenscanLine(lf, line);
slAddHead(&gsfList, gsf);
}
slReverse(&gsfList);
printf("Got %d exons\n", slCount(gsfList));
seg->geneList = bundleGenes(gsfList);
seg->geneList = filterEmptyGenes(seg->geneList);
gsfList = NULL;
printf("Got %d genes\n", slCount(seg->geneList));
if (!lineFileNext(lf, &line, NULL))
errAbort("Unexpected end of file in %s", lf->fileName);
if (startsWith("Suboptimal exons", line))
{
mustSkipTo(lf, "-----");
for (;;)
{
if (!lineFileNext(lf, &line, NULL))
break;
line = skipLeadingSpaces(line);
if (line == NULL || line[0] == 0)
continue;
if (!startsWith("S.", line))
break;
gsf = parseGenscanLine(lf, line);
slAddHead(&gsfList, gsf);
}
slReverse(&gsfList);
seg->suboptList = gsfList;
printf("Got %d suboptimal exons\n", slCount(seg->suboptList));
}
lineFileReuse(lf);
mustSkipTo(lf, "Predicted peptide sequence");
if ((line = skipTo(lf, ">")) != NULL)
{
lineFileReuse(lf);
for (gsg = seg->geneList; gsg != NULL; gsg = gsg->next)
{
aaSeq seq;
if (!faPepSpeedReadNext(lf, &seq.dna, &seq.size, &seq.name))
errAbort("Not enough predicted peptides in %s\n", lf->fileName);
gsg->translation = cloneString(seq.dna);
}
}
lineFileClose(&lf);
return seg;
}
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);
}
