void writeAgpFile(char *chromName, struct agpData *startAgpData, char *filename)
/*
Writes the agp file out for a single chromsome
param chromName - The name of the chromsome.
param startGap - Pointer to the dna gap or fragment at which we are starting to
write data. The data will include the contents of this gap/frag.
param filename - The file to write.
*/
{
struct agpData *curData = NULL;
FILE *fp = NULL;
fp = fopen(filename, "w");
/*printf("Writing agp file %s for chromo %s\n", filename, startAgpData->data.pGap->chrom);*/
curData = startAgpData;
while (NULL != curData)
{
if (curData->isGap)
{
/* Undo the decrement we did earlier.
This was done in nextAgpEntryToSplitOn() in order
to be compatible with the 0-based frag addressing scheme.
*/
curData->data.pGap->chromStart++;
agpGapOutput(curData->data.pGap, fp, '\t', '\n');
/* Redo the above decrement - don't want to create side effects */
curData->data.pGap->chromStart--;
}
else
{
agpFragOutput(curData->data.pFrag, fp, '\t', '\n');
}
curData = curData->next;
}
fclose(fp);
}
void scaffoldFaToAgp(char *scaffoldFile)
/* scaffoldFaToAgp - create AGP file, gap file and lift file
* from scaffold FA file */
{
DNA *scaffoldSeq;
char *name;
int size;
struct agpFrag frag;
struct agpGap scaffoldGap, fragGap;
struct lineFile *lf = lineFileOpen(scaffoldFile, TRUE);
char outDir[256], outFile[128], ext[64], outPath[512];
FILE *agpFile = NULL, *gapFile = NULL, *liftFile = NULL;
int fileNumber = 1; /* sequence number in AGP file */
int start = 0, end = 0;
int chromSize = 0;
int scaffoldCount = 0;
int fragSize = 0, gapSize = 0;
char *seq;
int seqStart = 0;
/* determine size of "unordered chromosome" that will be constructed.
* This is needed for the lift file. */
while (faMixedSpeedReadNext(lf, &scaffoldSeq, &size, &name))
{
chromSize += size;
chromSize += scaffoldGapSize;
scaffoldCount++;
}
/* do not need the final useless gap */
chromSize -= scaffoldGapSize;
printf("scaffold gap size is %d, total scaffolds: %d\n",
scaffoldGapSize, scaffoldCount);
printf("chrom size is %d\n", chromSize);
/* initialize fixed fields in AGP frag */
ZeroVar(&frag);
frag.chrom = CHROM_NAME;
frag.type[0] = 'D'; /* draft */
frag.fragStart = 0; /* always start at beginning of scaffold */
frag.strand[0] = '+';
/* initialize fixed fields in scaffold gap */
ZeroVar(&scaffoldGap);
scaffoldGap.chrom = CHROM_NAME;
scaffoldGap.n[0] = 'N';
scaffoldGap.size = scaffoldGapSize;
scaffoldGap.type = SCAFFOLD_GAP_TYPE;
scaffoldGap.bridge = "no";
/* initialize fixed fields in frag gap */
ZeroVar(&fragGap);
fragGap.chrom = CHROM_NAME;
fragGap.n[0] = 'N';
fragGap.type = FRAGMENT_GAP_TYPE;
fragGap.bridge = "yes";
/* munge file paths */
splitPath(scaffoldFile, outDir, outFile, ext);
sprintf(outPath, "%s%s.agp", outDir, outFile);
agpFile = mustOpen(outPath, "w");
printf("writing %s\n", outPath);
sprintf(outPath, "%s%s.gap", outDir, outFile);
gapFile = mustOpen(outPath, "w");
printf("writing %s\n", outPath);
sprintf(outPath, "%s%s.lft", outDir, outFile);
liftFile = mustOpen(outPath, "w");
printf("writing %s\n", outPath);
/* read in scaffolds from fasta file, and generate
* the three files */
lineFileSeek(lf, 0, SEEK_SET);
boolean allDone = FALSE;
allDone = ! faMixedSpeedReadNext(lf, &scaffoldSeq, &size, &name);
while (! allDone)
{
end = start + size;
/* setup AGP frag for the scaffold and write to AGP file */
frag.frag = name;
frag.ix = fileNumber++;
frag.chromStart = start;
frag.chromEnd = end;
frag.fragEnd = size;
agpFragOutput(&frag, agpFile, '\t', '\n');
/* write lift file entry for this scaffold */
fprintf(liftFile, "%d\t%s\t%d\t%s\t%d\n",
start, name, size, CHROM_NAME, chromSize);
/* write gap file entries for this scaffold */
seq = scaffoldSeq;
seqStart = start;
while (seqGetGap(seq, &fragSize, &gapSize))
{
if (gapSize > minGapSize)
{
fragGap.size = gapSize;
fragGap.chromStart = seqStart + fragSize + 1;
fragGap.chromEnd = fragGap.chromStart + gapSize - 1;
agpGapOutput(&fragGap, gapFile, '\t', '\n');
}
seqStart = seqStart + fragSize + gapSize;
seq = seq + fragSize + gapSize;
}
/* setup AGP gap to separate scaffolds and write to AGP and gap files */
/* Note: may want to suppress final gap -- not needed as separator */
start = end + 1;
end = start + scaffoldGapSize - 1;
/* Avoid an extra gap on the end - not needed */
allDone = ! faMixedSpeedReadNext(lf, &scaffoldSeq, &size, &name);
if (allDone)
break;
scaffoldGap.ix = fileNumber++;
scaffoldGap.chromStart = start;
scaffoldGap.chromEnd = end;
agpGapOutput(&scaffoldGap, agpFile, '\t', '\n');
agpGapOutput(&scaffoldGap, gapFile, '\t', '\n');
/* write lift file entry for this gap */
fprintf(liftFile, "%d\t%s\t%d\t%s\t%d\n",
start-1, "gap", scaffoldGapSize, CHROM_NAME, chromSize);
start = end;
//freeMem(seq);
}
carefulClose(&agpFile);
carefulClose(&liftFile);
carefulClose(&gapFile);
lineFileClose(&lf);
}
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);
}
