static void goldLoad(struct track *tg)
/* Load up golden path from database table to track items. */
{
struct sqlConnection *conn = hAllocConn(database);
struct sqlResult *sr = NULL;
char **row;
struct agpFrag *fragList = NULL, *frag;
struct agpGap *gapList = NULL, *gap;
int rowOffset;
/* Get the frags and load into tg->items. */
sr = hRangeQuery(conn, "gold", chromName, winStart, winEnd, NULL, &rowOffset);
while ((row = sqlNextRow(sr)) != NULL)
{
frag = agpFragLoad(row+rowOffset);
slAddHead(&fragList, frag);
}
slSort(&fragList, cmpAgpFrag);
sqlFreeResult(&sr);
tg->items = fragList;
/* Get the gaps into tg->customPt. */
sr = hRangeQuery(conn, "gap", chromName, winStart, winEnd, NULL, &rowOffset);
while ((row = sqlNextRow(sr)) != NULL)
{
gap = agpGapLoad(row+rowOffset);
slAddHead(&gapList, gap);
}
slReverse(&gapList);
sqlFreeResult(&sr);
tg->customPt = gapList;
hFreeConn(&conn);
}
struct agp *agpLoad(char **row, int ct)
/* Load an AGP entry from array of strings. Dispose with agpFree */
{
struct agp *agp;
struct agpFrag *agpFrag;
struct agpGap *agpGap;
if (ct < 8)
errAbort("Expecting >= 8 words in AGP file, got %d\n", ct);
AllocVar(agp);
if (row[4][0] != 'N' && row[4][0] != 'U')
{
/* not a gap */
if (ct != 9)
errAbort("Expecting 9 words in AGP fragment line, got %d\n", ct);
agpFrag = agpFragLoad(row);
agp->entry = agpFrag;
agp->isFrag = TRUE;
}
else
{
/* gap */
agpGap = agpGapLoad(row);
agp->entry = agpGap;
agp->isFrag = FALSE;
}
return agp;
}
static void agpToFa(char *agpFile, char *agpSeq, char *faOut, char *seqDir)
/* agpToFa - Convert a .agp file to a .fa file. */
{
struct lineFile *lf = lineFileOpen(agpFile, TRUE);
char *line, *words[16];
int lineSize, wordCount;
int lastPos = 0;
struct agpFrag *agpList = NULL, *agp;
FILE *f = mustOpen(faOut, "w");
char *prevChrom = NULL;
verbose(2,"#\tprocessing AGP file: %s\n", agpFile);
while (lineFileNext(lf, &line, &lineSize))
{
if (line[0] == 0 || line[0] == '#' || line[0] == '\n')
continue;
wordCount = chopLine(line, words);
if (wordCount < 5)
errAbort("Bad line %d of %s: need at least 5 words, got %d\n",
lf->lineIx, lf->fileName, wordCount);
if (! (sameWord("all", agpSeq) || sameWord(words[0], agpSeq)))
continue;
if (prevChrom != NULL && !sameString(prevChrom, words[0]))
{
agpToFaOne(&agpList, agpFile, prevChrom, seqDir, lastPos, f);
lastPos = 0;
}
if (words[4][0] != 'N' && words[4][0] != 'U')
{
lineFileExpectAtLeast(lf, 9, wordCount);
agp = agpFragLoad(words);
/* file is 1-based but agpFragLoad() now assumes 0-based: */
agp->chromStart -= 1;
agp->fragStart -= 1;
if (agp->chromStart != lastPos)
errAbort("Start doesn't match previous end line %d of %s\n",
lf->lineIx, lf->fileName);
if (agp->chromEnd - agp->chromStart != agp->fragEnd - agp->fragStart)
errAbort("Sizes don't match in %s and %s line %d of %s\n",
agp->chrom, agp->frag, lf->lineIx, lf->fileName);
slAddHead(&agpList, agp);
lastPos = agp->chromEnd;
}
else
{
lastPos = lineFileNeedNum(lf, words, 2);
}
if (prevChrom == NULL || !sameString(prevChrom, words[0]))
{
freeMem(prevChrom);
prevChrom = cloneString(words[0]);
}
}
agpToFaOne(&agpList, agpFile, prevChrom, seqDir, lastPos, f);
}
struct scaffold *readScaffoldsFromAgp(char *fileName)
/* Read in agp file and return as list of scaffolds. */
{
struct hash *scaffoldHash = newHash(17);
struct lineFile *lf = lineFileOpen(fileName, TRUE);
char *row[9];
int wordCount;
struct scaffold *scaffoldList = NULL, *scaffold;
struct agpFrag *frag;
int size;
for (;;)
{
wordCount = lineFileChop(lf, row);
if (wordCount <= 0)
break;
if (wordCount < 8)
lineFileShort(lf);
if (row[4][0] == 'N' || row[4][0] == 'U')
continue;
if (wordCount < 9)
lineFileShort(lf);
frag = agpFragLoad(row);
frag->chromStart -= 1;
frag->fragStart -= 1;
size = frag->fragEnd - frag->fragStart;
if (size != frag->chromEnd - frag->chromStart)
errAbort("scaffold/contig size mismatch line %d of %s", lf->lineIx, lf->fileName);
if (frag->strand[0] != '+')
errAbort("Strand not + line %d of %s", lf->lineIx, lf->fileName);
scaffold = hashFindVal(scaffoldHash, frag->chrom);
if (scaffold == NULL)
{
AllocVar(scaffold);
hashAdd(scaffoldHash, frag->chrom, scaffold);
slAddHead(&scaffoldList, scaffold);
}
slAddHead(&scaffold->list, frag);
if (frag->chromEnd > scaffold->size)
scaffold->size = frag->chromEnd;
}
slReverse(&scaffoldList);
for (scaffold = scaffoldList; scaffold != NULL; scaffold = scaffold->next)
slReverse(&scaffold->list);
printf("Got %d scaffolds in %s\n", slCount(scaffoldList), lf->fileName);
lineFileClose(&lf);
hashFree(&scaffoldHash);
return scaffoldList;
}
struct agpFrag *loadChromAgp(struct sqlConnection *conn, char *chrom)
/* Load all AGP fragments for chromosome. */
{
char query[256];
struct sqlResult *sr;
char **row;
struct agpFrag *fragList = NULL, *frag;
sqlSafef(query, sizeof query, "select * from %s_gold", chrom);
sr = sqlGetResult(conn, query);
while ((row = sqlNextRow(sr)) != NULL)
{
frag = agpFragLoad(row);
slAddHead(&fragList, frag);
}
slReverse(&fragList);
return fragList;
}
struct agpFrag *readAgpFile(char *agpName)
/* Read agps from file. */
{
struct lineFile *lf = lineFileOpen(agpName, TRUE);
int wordCount;
char *words[16];
struct agpFrag *list = NULL, *el;
while ((wordCount = lineFileChop(lf, words)) != 0)
{
if (words[4][0] != 'N')
{
lineFileExpectWords(lf, 9, wordCount);
el = agpFragLoad(words);
slAddHead(&list, el);
}
}
lineFileClose(&lf);
slReverse(&list);
return list;
}
struct agpData* nextAgpEntryToSplitOn(struct lineFile *lfAgpFile, int dnaSize, struct agpData **retStartData)
/*
Finds the next agp entry in the agp file at which to split on.
param lfAgpFile - The .agp file we are examining.
param dnaSize - The total size of the chromsome's dna sequence
we are splitting up. Used to prevent overrun of the algorithm
that looks at agp entries.
param retStartData - An out param returning the starting(inclusive) gap that we
will start to split on.
return struct agpData* - The ending (inclusive) agp data we are to split on.
*/
{
int startIndex = 0;
int numBasesRead = 0;
char *line = NULL;
char *words[9];
int lineSize = 0;
struct agpGap *agpGap = NULL;
struct agpFrag *agpFrag = NULL;
struct agpData *curAgpData = NULL;
struct agpData *prevAgpData = NULL;
boolean splitPointFound = FALSE;
int splitSize = _nSize;
do
{
lineFileNext(lfAgpFile, &line, &lineSize);
if (line[0] == '#' || line[0] == '\n')
{
continue;
}
AllocVar(curAgpData);
curAgpData->endOfContig = FALSE;
curAgpData->isGap = FALSE;
curAgpData->prev = NULL;
curAgpData->next = NULL;
chopLine(line, words);
if (words[4][0] == 'N' || words[4][0] == 'U')
{
agpGap = agpGapLoad(words);
/*
Decrement the chromStart index since that's how the agpFrags do it
and we want to use 0-based addressing
*/
--(agpGap->chromStart);
if (0 == startIndex)
{
startIndex = agpGap->chromStart;
}
if (numBasesRead >= _bSize)
{
splitPointFound = TRUE;
}
else
{
/* Split points are made after non-bridged contigs only here */
splitPointFound = (0 == strcasecmp(agpGap->bridge, NO));
}
curAgpData->isGap = TRUE;
curAgpData->data.pGap = agpGap;
}
else
{
agpFrag = agpFragLoad(words);
// file is 1-based but agpFragLoad() now assumes 0-based:
agpFrag->chromStart -= 1;
agpFrag->fragStart -= 1;
/* If we find a fragment and not a gap */
if (0 == startIndex)
{
startIndex = agpFrag->chromStart;
}
if (numBasesRead >= _aSize)
{
splitPointFound = TRUE;
}
else
{
splitPointFound = FALSE;
}
curAgpData->isGap = FALSE;
curAgpData->data.pFrag = agpFrag;
}
/* Since this our first loop iteration,
save the start gap as the beginning of the section to write out */
if (NULL == prevAgpData)
{
*retStartData = curAgpData; /* Save the pointer to the head of the list */
}
else
{
/* Build a doubly linked list for use elsewhere */
prevAgpData->next = curAgpData;
curAgpData->prev = prevAgpData;
}
prevAgpData = curAgpData;
numBasesRead = curAgpData->data.pGap->chromEnd - startIndex;
} while ((numBasesRead < splitSize || !splitPointFound)
&& curAgpData->data.pGap->chromEnd < dnaSize);
curAgpData->next = NULL; /* Terminate the linked list */
curAgpData->endOfContig = TRUE;
return curAgpData;
}
static void agpSangerUnfinished(char *agpFile, char *contigFasta, char *agpOut)
/* Fix agp to match unfinished contigs in fasta */
{
struct lineFile *lf = lineFileOpen(agpFile, TRUE);
char *line, *words[16];
int lineSize, wordCount;
unsigned lastPos = 0;
struct agpFrag *agp;
struct agpGap *gap;
FILE *f;
char *lastObj = NULL;
f = mustOpen(agpOut, "w");
char *newChrom = NULL;
struct hash *hash = hashFasta(contigFasta);
verbose(2,"#\tprocessing AGP file: %s\n", agpFile);
while (lineFileNext(lf, &line, &lineSize))
{
if (line[0] == 0 || line[0] == '#' || line[0] == '\n')
continue;
//verbose(2,"#\tline: %d\n", lf->lineIx);
wordCount = chopLine(line, words);
if (wordCount < 5)
errAbort("Bad line %d of %s: need at least 5 words, got %d\n",
lf->lineIx, lf->fileName, wordCount);
if (!lastObj || !sameString(words[0],lastObj))
{
freez(&newChrom);
newChrom = cloneString(words[0]);
lastPos = 0;
}
if (words[4][0] != 'N')
{
lineFileExpectAtLeast(lf, 9, wordCount);
agp = agpFragLoad(words);
/* agp is 1-based but agp loaders do not adjust for 0-based: */
agp->chromStart -= 1;
agp->fragStart -= 1;
if (agp->chromEnd - agp->chromStart != agp->fragEnd - agp->fragStart)
errAbort("Sizes don't match in %s and %s line %d of %s\n",
agp->chrom, agp->frag, lf->lineIx, lf->fileName);
char *root = cloneString(agp->frag);
chopSuffixAt(root, '.');
struct hashEl *e, *elist = hashLookup(hash, root);
for (e = elist; e; e = hashLookupNext(e))
{
struct unfinishedContig *u = e->val;
if ((u->fragStart <= agp->fragStart) && (u->fragEnd >= agp->fragEnd))
{
agp->frag = cloneString(u->frag);
agp->fragEnd -= u->fragStart;
agp->fragStart -= u->fragStart;
}
}
freeMem(root);
}
else
{
lineFileExpectAtLeast(lf, 8, wordCount);
gap = agpGapLoad(words);
/* to be consistent with agpFrag */
gap->chromStart -= 1;
agp = (struct agpFrag*)gap;
}
if (agp->chromStart != lastPos)
errAbort("Start doesn't match previous end line %d of %s\n"
"agp->chromStart: %u\n"
"agp->chromEnd: %u\n"
"lastPos: %u\n"
,lf->lineIx, lf->fileName
,agp->chromStart
,agp->chromEnd
,lastPos
);
lastPos = agp->chromEnd;
freez(&lastObj);
lastObj = cloneString(words[0]); /* not agp->chrom which may be modified already */
if (words[4][0] != 'N')
{
/* agpFragOutput assumes 0-based-half-open, but writes 1-based for agp */
agpFragOutput(agp, f, '\t', '\n');
agpFragFree(&agp);
}
else
{
/* restore back to 1-based for agp
* because agpGapOutput doesn't compensate */
gap->chromStart += 1;
agpGapOutput(gap, f, '\t', '\n');
agpGapFree(&gap);
}
}
carefulClose(&f);
}
static void agpMergeChromScaf(char *agpFile, char *agpOut, boolean filtering)
/* Create a combined agp file from the chrom.agp and scaffold.agp,
* merging in only scaffolds from scaffold.agp
* that are not already in chroms. */
{
struct lineFile *lf = lineFileOpen(agpFile, TRUE);
char *line, *words[16];
int lineSize, wordCount;
unsigned lastPos = 0;
struct agpFrag *agp;
struct agpGap *gap;
FILE *f;
char *lastObj = NULL;
f = mustOpen(agpOut, filtering ? "a" : "w");
char *newChrom = NULL;
static struct hash *hash = NULL;
boolean skipping = FALSE;
if (!hash)
hash = hashNew(0);
verbose(2,"#\tprocessing AGP file: %s\n", agpFile);
while (lineFileNext(lf, &line, &lineSize))
{
if (line[0] == 0 || line[0] == '#' || line[0] == '\n')
continue;
//verbose(2,"#\tline: %d\n", lf->lineIx);
wordCount = chopLine(line, words);
if (wordCount < 5)
errAbort("Bad line %d of %s: need at least 5 words, got %d\n",
lf->lineIx, lf->fileName, wordCount);
if (!lastObj || !sameString(words[0],lastObj))
{
freez(&newChrom);
newChrom = cloneString(words[0]);
lastPos = 0;
}
skipping = FALSE;
if (filtering)
{
if (hashLookup(hash, words[0]))
skipping = TRUE;
}
if (words[4][0] != 'N')
{
lineFileExpectAtLeast(lf, 9, wordCount);
agp = agpFragLoad(words);
/* agp is 1-based but agp loaders do not adjust for 0-based: */
agp->chromStart -= 1;
agp->fragStart -= 1;
if (agp->chromEnd - agp->chromStart != agp->fragEnd - agp->fragStart)
errAbort("Sizes don't match in %s and %s line %d of %s\n",
agp->chrom, agp->frag, lf->lineIx, lf->fileName);
if (!filtering)
{
char *root = cloneString(agp->frag);
chopSuffixAt(root, '.');
hashStore(hash, root);
freeMem(root);
}
}
else
{
lineFileExpectAtLeast(lf, 8, wordCount);
gap = agpGapLoad(words);
/* to be consistent with agpFrag */
gap->chromStart -= 1;
agp = (struct agpFrag*)gap;
}
if (agp->chromStart != lastPos)
errAbort("Start doesn't match previous end line %d of %s\n"
"agp->chromStart: %u\n"
"agp->chromEnd: %u\n"
"lastPos: %u\n"
,lf->lineIx, lf->fileName
,agp->chromStart
,agp->chromEnd
,lastPos
);
lastPos = agp->chromEnd;
freez(&lastObj);
lastObj = cloneString(words[0]); /* not agp->chrom which may be modified already */
if (words[4][0] != 'N')
{
/* agpFragOutput assumes 0-based-half-open, but writes 1-based for agp */
if (!skipping)
agpFragOutput(agp, f, '\t', '\n');
agpFragFree(&agp);
}
else
{
/* restore back to 1-based for agp
* because agpGapOutput doesn't compensate */
gap->chromStart += 1;
if (!skipping)
agpGapOutput(gap, f, '\t', '\n');
agpGapFree(&gap);
}
}
carefulClose(&f);
}
struct hash *agpLoadAll(char *agpFile)
/* load AGP entries into a hash of AGP lists, one per chromosome */
{
struct hash *agpHash = newHash(0);
struct lineFile *lf = lineFileOpen(agpFile, TRUE);
char *words[9];
int lastPos = 0;
int wordCount;
struct agpFrag *agpFrag;
struct agpGap *agpGap;
char *chrom;
struct agp *agp;
struct hashEl *hel;
while ((wordCount = lineFileChopNext(lf, words, ArraySize(words))) != 0)
{
lineFileExpectAtLeast(lf, 8, wordCount);
chrom = words[0];
if (!hashFindVal(agpHash, chrom))
lastPos = 1;
AllocVar(agp);
if (words[4][0] != 'N' && words[4][0] != 'U')
{
/* not a gap */
lineFileExpectWords(lf, 9, wordCount);
agpFrag = agpFragLoad(words);
if (agpFrag->chromStart != lastPos)
errAbort(
"Frag start (%d, %d) doesn't match previous end line %d of %s\n",
agpFrag->chromStart, lastPos, lf->lineIx, lf->fileName);
if (agpFrag->chromEnd - agpFrag->chromStart !=
agpFrag->fragEnd - agpFrag->fragStart)
errAbort("Sizes don't match in %s and %s line %d of %s\n",
agpFrag->chrom, agpFrag->frag, lf->lineIx, lf->fileName);
lastPos = agpFrag->chromEnd + 1;
agp->entry = agpFrag;
agp->isFrag = TRUE;
}
else
{
/* gap */
lineFileExpectWords(lf, 8, wordCount);
agpGap = agpGapLoad(words);
if (agpGap->chromStart != lastPos)
errAbort("Gap start (%d, %d) doesn't match previous end line %d of %s\n",
agpGap->chromStart, lastPos, lf->lineIx, lf->fileName);
lastPos = agpGap->chromEnd + 1;
agp->entry = agpGap;
agp->isFrag = FALSE;
}
if ((hel = hashLookup(agpHash, chrom)) == NULL)
hashAdd(agpHash, chrom, agp);
else
slAddHead(&(hel->val), agp);
}
#ifndef DEBUG
{
struct hashCookie cookie;
struct hashEl *hel;
cookie = hashFirst(agpHash);
while ((hel = hashNext(&cookie)) != NULL)
{
struct agp *agpList;
agpList = (struct agp *)hel->val;
/*
for (agp = agpList; agp != NULL; agp = agp->next)
printf("isFrag: %d\n", agp->isFrag);
*/
}
}
#endif
/* reverse AGP lists */
//hashTraverseVals(agpHash, slReverse);
#ifndef DEBUG
{
struct hashCookie cookie;
struct hashEl *hel;
cookie = hashFirst(agpHash);
while ((hel = hashNext(&cookie)) != NULL)
{
struct agp *agpList;
slReverse(&hel->val);
agpList = hel->val;
/*
agpList = (struct agp *)hel->val;
slReverse(&agpList);
hashRemove(agpHash, hel->name);
hashAdd(agpHash, hel->name, agpList);
*/
/*
for (agp = agpList; agp != NULL; agp = agp->next)
printf("isFrag: %d\n", agp->isFrag);
*/
}
}
#endif
return agpHash;
}
void fakeFinContigs(char *agpName, char *faName, char *finDir, char *rootName, char *finFaDir, char *ooVer)
/* fakeFinContigs - Fake up contigs for a finished chromosome. */
{
struct contig *contigList = NULL, *contig = NULL;
struct agpFrag *agp;
struct lineFile *lf = lineFileOpen(agpName, TRUE);
char *line, *words[16];
int lineSize, wordCount;
int contigIx = 0;
char liftDir[512], contigDir[512], path[512];
char chrom[128];
FILE *f;
struct dnaSeq *seq;
int fragIx;
/* Build up contig list by scanning agp file. */
printf("Reading %s\n", lf->fileName);
while (lineFileNext(lf, &line, &lineSize))
{
if (line[0] == '#' || line[0] == 0)
continue;
wordCount = chopLine(line, words);
if (wordCount < 5)
errAbort("Expecting at least 5 words line %d of %s", lf->lineIx, lf->fileName);
if (words[4][0] == 'N' || words[4][0] == 'U')
{
contig = NULL;
continue;
}
lineFileExpectWords(lf, 9, wordCount);
agp = agpFragLoad(words);
// file is 1-based but agpFragLoad() now assumes 0-based:
agp->chromStart -= 1;
agp->fragStart -= 1;
if (contig == NULL)
{
AllocVar(contig);
sprintf(contig->name, "%s%d", rootName, ++contigIx);
contig->startOffset = agp->chromStart;
slAddHead(&contigList, contig);
}
else
{
if (contig->agpList != NULL && contig->agpList->chromEnd != agp->chromStart)
errAbort("Start doesn't match previous end line %d of %s",
lf->lineIx, lf->fileName);
}
if (agp->chromEnd - agp->chromStart != agp->fragEnd - agp->fragStart)
errAbort("Chrom and frag size mismatch line %d of %s", lf->lineIx, lf->fileName);
slAddHead(&contig->agpList, agp);
contig->endOffset = agp->chromEnd;
}
slReverse(&contigList);
for (contig = contigList; contig != NULL; contig = contig->next)
slReverse(&contig->agpList);
lineFileClose(&lf);
/* Load up chromosome sequence and make sure it is in one piece. */
printf("Reading %s\n", faName);
seq = faReadAllDna(faName);
if (slCount(seq) != 1)
errAbort("Got %d sequences in %s, can only handle one.", slCount(seq), faName);
/* Fix up agp coordinates. Make a directory for each contig. Fill it with
* .fa .agp barge.NN files for that contig. */
printf("Writing contig dirs\n");
for (contig = contigList; contig != NULL; contig = contig->next)
{
/* Make Contig dir. */
sprintf(contigDir, "%s/%s", finDir, contig->name);
makeDir(contigDir);
/* Make contig.agp file. */
sprintf(path, "%s/%s.agp", contigDir, contig->name);
f = mustOpen(path, "w");
fragIx = 0;
for (agp = contig->agpList; agp != NULL; agp = agp->next)
{
char buf[128];
sprintf(buf, "%s/%s", skipChr(agp->chrom), contig->name);
freez(&agp->chrom);
agp->chrom = cloneString(buf);
agp->chromStart -= contig->startOffset;
agp->chromEnd -= contig->startOffset;
agp->ix = ++fragIx;
agpFragTabOut(agp, f);
}
carefulClose(&f);
/* Make ooGreedy.NN.gl file */
sprintf(path, "%s/%s.%s.gl", contigDir, "ooGreedy", ooVer);
f = mustOpen(path, "w");
for (agp = contig->agpList; agp != NULL; agp = agp->next)
{
if (agp->type[0] != 'N' && agp->type[0] != 'U')
{
fprintf(f, "%s_1\t%d\t%d\t%s\n", agp->frag,
agp->chromStart,
agp->chromEnd,
agp->strand);
}
}
carefulClose(&f);
/* Make contig.fa file. */
sprintf(path, "%s/%s.fa", contigDir, contig->name);
faWrite(path, contig->name, seq->dna + contig->startOffset,
contig->endOffset - contig->startOffset);
/* Make contig/barge file. */
sprintf(path, "%s/barge.%s", contigDir, ooVer);
f = mustOpen(path, "w");
fprintf(f, "Barge (Connected Clone) File ooGreedy Version %s\n", ooVer);
fprintf(f, "\n");
fprintf(f, "start accession size overlap maxClone maxOverlap\n");
fprintf(f, "------------------------------------------------------------\n");
for (agp = contig->agpList; agp != NULL; agp = agp->next)
{
char clone[128];
strcpy(clone, agp->frag);
chopSuffix(clone);
fprintf(f, "%d\t%s\t%d\t100\tn/a\t0\n", agp->chromStart,
clone, agp->chromEnd);
}
carefulClose(&f);
/* Make contig/gold file. */
sprintf(path, "%s/gold.%s", contigDir, ooVer);
f = mustOpen(path, "w");
fragIx = 0;
for (agp = contig->agpList; agp != NULL; agp = agp->next)
{
char fragName[128];
struct agpFrag frag = *agp;
sprintf(fragName, "%s_1", agp->frag);
frag.frag = fragName;
frag.type[0] = '0';
agpFragTabOut(&frag, f);
}
carefulClose(&f);
}
/* Create lift subdirectory. */
printf("Creating lift files\n");
sprintf(liftDir, "%s/lift", finDir);
makeDir(liftDir);
/* Create lift/oOut.lst file (just a list of contigs). */
sprintf(path, "%s/oOut.lst", liftDir);
f = mustOpen(path, "w");
for (contig = contigList; contig != NULL; contig = contig->next)
fprintf(f, "%s/%s.fa.out\n", contig->name, contig->name);
carefulClose(&f);
/* Create lift/ordered.lst file (just a list of contigs). */
sprintf(path, "%s/ordered.lst", liftDir);
f = mustOpen(path, "w");
for (contig = contigList; contig != NULL; contig = contig->next)
fprintf(f, "%s\n", contig->name);
carefulClose(&f);
/* Create lift/ordered.lft file. */
sprintf(path, "%s/ordered.lft", liftDir);
f = mustOpen(path, "w");
splitPath(faName, NULL, chrom, NULL);
for (contig = contigList; contig != NULL; contig = contig->next)
fprintf(f, "%d\t%s/%s\t%d\t%s\t%d\n",
contig->startOffset, skipChr(chrom), contig->name,
contig->endOffset - contig->startOffset,
chrom, seq->size);
carefulClose(&f);
}
struct chrom *readChromScaffoldsFromAgp(char *fileName)
/* Read in agp file and return as list of chroms. */
{
struct hash *chromHash = newHash(17);
struct lineFile *lf = lineFileOpen(fileName, TRUE);
char *row[9];
int wordCount;
struct chrom *chromList = NULL, *chrom;
struct agpFrag *frag = NULL;
struct agpGap *gap = NULL;
char *chromName;
int chromSize = 0;
for (;;)
{
wordCount = lineFileChop(lf, row);
if (wordCount <= 0)
break;
if (wordCount < 8)
lineFileShort(lf);
if (row[4][0] == 'N' || row[4][0] == 'U')
{
/* need to get chromEnd from gaps to determine chrom size
* if the chrom ends with a gap */
gap = agpGapLoad(row);
chromName = gap->chrom;
chromSize = gap->chromEnd;
frag = NULL;
}
else
{
if (wordCount < 9)
lineFileShort(lf);
frag = agpFragLoad(row);
chromName = frag->chrom;
chromSize = frag->chromEnd;
frag->chromStart -= 1;
frag->fragStart -= 1;
if (frag->fragEnd - frag->fragStart !=
frag->chromEnd - frag->chromStart)
errAbort("chrom/scaffold size mismatch line %d of %s",
lf->lineIx, lf->fileName);
}
chrom = hashFindVal(chromHash, chromName);
if (chrom == NULL)
{
AllocVar(chrom);
slAddHead(&chromList, chrom);
hashAdd(chromHash, chromName, chrom);
}
chrom->size = max(chromSize, chrom->size);
if (frag != NULL)
slAddHead(&chrom->list, frag);
}
slReverse(&chromList);
for (chrom = chromList; chrom != NULL; chrom = chrom->next)
slReverse(&chrom->list);
verbose(1, "Got %d chroms in %s\n", slCount(chromList), lf->fileName);
lineFileClose(&lf);
hashFree(&chromHash);
return chromList;
}
