static void hackSynClone()
/* Make edits to synthetic clone entries that appear to be mRNAs to have a mol type of
* mRNA (normally marked as DNA) */
{
if (kvtGet(kvt, "gen") != NULL)
{
struct keyVal *kv = kvtGet(kvt, "mol");
if (kv != NULL)
kv->val = "mRNA";
}
}
char *kvtLookup(struct kvt *kvt, char *key)
/* Search table for key. Return key value, or NULL if
* key not found. */
{
struct keyVal *keyVal = kvtGet(kvt, key);
if (keyVal == NULL)
return NULL;
else
return keyVal->val;
}
static boolean keepGbEntry(boolean isEst)
/* should the current entry in the kvt be kept? */
{
char *acc = gbAccessionField->val->string;
char *cat = kvtGet(kvt, "cat")->val;
char *org = kvtGet(kvt, "org")->val;
if (genbankBlackListFail(acc, blackListRanges))
return FALSE;
else if (gbOrg && (differentString(org, gbOrg)))
return FALSE;
else if (gbGuessSrcDb(acc) == GB_REFSEQ)
{
return (startsWith("NM_", acc) || startsWith("NR_", acc)
|| ((startsWith("XM_", acc) && inclXMs)));
}
else if ((sameString(cat, "GSS") ||
sameString(cat, "HTG") ||
sameString(cat, "STS") ||
sameString(cat, "CON")) && !(gbType & GB_DNA))
return FALSE; // division to ignore
else
{
if (sameString(cat, "EST")) {
return (gbType & GB_EST) != 0;
}
else if (gbType & GB_MRNA)
{
// not an EST, keep any type of RNA
return containsStringNoCase(kvtGet(kvt, "mol")->val, "RNA") != NULL;
}
else if (gbType & GB_DNA)
{
// keep any type of DNA if DNA filter set
return containsStringNoCase(kvtGet(kvt, "mol")->val, "DNA") != NULL;
}
else
return FALSE;
}
}
static char *findSyntheticTarget()
/* for a synthetic sequence, attempt to find the targeted organism. This was
* added to support the MGC/ORFeome clones. In general, there is no defined way to
* determine an organism that a synthenic clone targets. */
{
struct keyVal *kv;
if (synOrgBuf == NULL)
synOrgBuf = dyStringNew(256);
dyStringClear(synOrgBuf);
kv = kvtGet(kvt, "srcOrg");
if (kv != NULL)
dyStringAppend(synOrgBuf, kv->val);
if (synOrgBuf->stringSize > 0)
{
kvtAdd(kvt, "synOrg", synOrgBuf->string);
return synOrgBuf->string;
}
else
return NULL;
}
static void procGbEntry(struct lineFile *lf, struct hash *estAuthorHash)
/* process one entry in the genbank file . readGbInfo should be called
* first */
{
char *words[16];
char date[64];
int wordCount;
DNA *dna = NULL;
int dnaSize;
char sizeString[16];
char accVer[64];
int faSize;
char *locus = gbLocusField->val->string;
char *accession = gbAccessionField->val->string;
int version = 0;
char *gi = NULL;
char *verChar = gbVersionField->val->string;
char *s;
char *org = gbOrganismField->val->string;
char *synOrg = NULL;
struct keyVal *seqKey, *sizeKey, *commentKey;
boolean isEst = FALSE;
char verNum[8];
char *com = gbCommentField->val->string;
if (locus == NULL || accession == NULL)
errAbort("No LOCUS or no ACCESSION line near %d of %s",
lf->lineIx, lf->fileName);
lmCleanup(&kvtMem);
/* Chop off all but first word of accession. */
s = skipLeadingSpaces(accession);
if (s != NULL)
s = skipLeadingNonSpaces(s);
if (s != NULL)
*s = 0;
/* Get version field (defaults to zero) */
if (verChar != NULL)
{
char *parts[2];
char *accVer;
int partCount;
partCount = chopByWhite(verChar, parts, ArraySize(parts));
/* Version is number after dot. */
accVer = parts[0];
if ((accVer = strchr(accVer, '.')) != NULL)
version = atoi(accVer+1);
if (partCount >= 2 && startsWith("GI:", parts[1]))
gi = parts[1]+3;
}
gbfFlatten(kvt);
/* Get additional keys. */
if (com != NULL)
{
if (startsWith("REVIEWED", com))
kvtAdd(kvt, "cur", "yes");
}
safef(verNum, sizeof(verNum), "%d", version);
kvtAdd(kvt, "ver", verNum);
if (gi != NULL)
kvtAdd(kvt, "ngi", gi);
wordCount = chopLine(locus, words);
if (wordCount >= 6)
{
kvtAdd(kvt, "mol", words[3]);
kvtAdd(kvt, "cat", words[wordCount-2]);
ncbiDateToSqlDate(words[wordCount-1], date, sizeof(date), accession);
kvtAdd(kvt, "dat", date);
}
else if (wordCount == 5 && sameString(words[2], "bp") && isdigit(words[1][0]))
{
/* Check carefully. Probably it's just missing the molecule type... */
if (!isNcbiDate(words[4]))
{
errAbort("Strange LOCUS line in %s accession %s",
lf->fileName, accession);
}
kvtAdd(kvt, "cat", words[3]);
ncbiDateToSqlDate(words[4], date, sizeof(date), accession);
kvtAdd(kvt, "dat", date);
}
else if (wordCount == 5 && sameString(words[2], "bp") && isdigit(words[1][0]))
{
kvtAdd(kvt, "mol", words[3]);
}
else
{
errAbort("Short LOCUS line in %s accession %s",
lf->fileName, accession);
}
if (((wordCount >= 5) && sameString(words[4], "EST")) ||
((wordCount >= 6) && sameString(words[5], "EST")))
{
/* Try and figure out if it's a 3' or 5' EST */
char *dir = getEstDir(gbDefinitionField->val->string, com);
if (dir != NULL)
kvtAdd(kvt, "dir", dir);
isEst = TRUE;
}
/* Handle other fields */
parseDbXrefs();
parseGene();
parseSourceOrganism();
parseMiscDiffs();
parseWarnings();
if (startsWith("synthetic construct", gbOrganismField->val->string))
{
synOrg = findSyntheticTarget();
if (synOrg != NULL)
hackSynClone();
}
if (keepGbEntry(isEst))
{
/* Handle sequence part of read. */
dna = gbfReadSequence(lf, &dnaSize);
}
/* just discard if no sequence */
if (dna != NULL)
{
seqKey = kvtAdd(kvt, "seq", dna);
safef(sizeString, sizeof(sizeString), "%d", dnaSize);
sizeKey = kvtAdd(kvt, "siz", sizeString);
if (isEst)
{
char *author = gbAuthorsField->val->string;
if (author != NULL)
{
struct authorExample *ae;
struct hashEl *hel;
if ((hel = hashLookup(estAuthorHash, author)) == NULL)
{
AllocVar(ae);
hel = hashAdd(estAuthorHash, author, ae);
ae->name = hel->name;
ae->count = 1;
strncpy(ae->accession, accession, sizeof(ae->accession));
slAddHead(&estAuthorList, ae);
}
else
{
ae = hel->val;
ae->count += 1;
}
}
}
seqKey->val = NULL; /* Don't write out sequence here. */
commentKey = kvtGet(kvt, "com");
if (commentKey != NULL)
commentKey->val = NULL; /* Don't write out comment either. */
setupOutputFiles(accession, org);
if (faFile != NULL)
{
/* save fasta offset, size in ra */
safef(accVer, sizeof(accVer), "%s.%d", accession, version);
gbFaWriteSeq(faFile, accVer, NULL, dna, -1);
faSize = faFile->off - faFile->recOff;
safef(faOffStr, sizeof(faOffStr), "%lld", (long long)faFile->recOff);
kvtAdd(kvt, "fao", faOffStr);
safef(faSizeStr, sizeof(faSizeStr), "%d", faSize);
kvtAdd(kvt, "fas", faSizeStr);
}
if (gPepFa != NULL)
{
/* must write before writing kvt */
writePepSeq();
}
kvtWriteAll(kvt, raFile, NULL);
if (gbIdxFile != NULL)
{
/* use synthetic target if it was determined */
struct keyVal *molkv = kvtGet(kvt, "mol");
enum molType molType = (molkv->val != NULL) ? gbParseMolType(molkv->val) : mol_mRNA;
gbProcessedWriteIdxRec(gbIdxFile, accession, version,
kvtLookup(kvt, "dat"),
((synOrg != NULL) ? synOrg : org),
molType);
}
}
else
gbfSkipSequence(lf);
}
