void wormClipRangeToChrom(char *chrom, int *pStart, int *pEnd)
/* Make sure that we stay inside chromosome. */
{
int chromEnd = wormChromSize(chrom);
int temp;
/* Swap ends if reversed. */
if (*pStart > *pEnd)
{
temp = *pEnd;
*pEnd = *pStart;
*pStart = temp;
}
/* Generally speaking try to slide the range covered by
* start-end inside the chromosome rather than just
* truncating an end. */
if (*pStart < 0)
{
*pEnd -= *pStart;
*pStart = 0;
}
if (*pEnd > chromEnd)
{
*pStart -= *pEnd - chromEnd;
*pEnd = chromEnd;
}
/* This handles case where the range is larger than the chromosome. */
if (*pStart < 0)
*pStart = 0;
}
int main(int argc, char *argv[])
{
#define stepSize 10000
#define extraBases 1000
static struct noiseTrack noiseTrack[stepSize];
int chromIx;
int chromSize;
int baseOff;
char *chromName;
int dnaStart, dnaEnd;
char *outName;
FILE *out;
struct hash *dupeHash;
if (argc != 2)
{
errAbort("editbase - lists bases for which there is evidence of RNA editing\n"
"usage:\n"
" editbase outfile.txt");
}
dnaUtilOpen();
initVlookup();
outName = argv[1];
out = mustOpen(outName, "w");
printf("Scanning for cDNAs that align more than once.\n");
dupeHash = buildMultiAlignHash();
printf("Loading worm genome\n");
wormLoadNt4Genome(&chrom, &chromCount);
wormChromNames(&chromNames, &chromCount);
for (chromIx = 0; chromIx < chromCount; ++chromIx)
{
chromName = chromNames[chromIx];
printf("Processing chromosome %s\n", chromName);
chromSize = wormChromSize(chromName);
for (baseOff = 0; baseOff < chromSize; baseOff += stepSize)
{
struct wormFeature *cdnaNamesList, *name;
struct cdnaAli *caList = NULL, *ca;
int dnaSize;
DNA *dna;
int chunkSize;
DNA *chunk;
int i;
/* Figure out how much DNA to get and get it. Include some
* extra around chunk so can align better. */
chunkSize = chromSize - baseOff;
if (chunkSize > stepSize) chunkSize = stepSize;
dnaStart = baseOff - extraBases;
dnaEnd = baseOff + stepSize + extraBases;
wormClipRangeToChrom(chromName, &dnaStart, &dnaEnd);
dnaSize = dnaEnd - dnaStart;
dna = wormChromPart(chromName, dnaStart, dnaSize);
/* Get the cDNAs */
cdnaNamesList = wormCdnasInRange(chromName, baseOff, baseOff + chunkSize);
for (name = cdnaNamesList; name != NULL; name = name->next)
{
if (!hashLookup(dupeHash, name->name) )
{
ca = makeCdnaAli(name->name, dna, dnaSize);
slAddHead(&caList, ca);
}
}
slReverse(&caList);
/* Add cdnas to noise track. */
chunk = dna + baseOff - dnaStart;
for (ca = caList; ca != NULL; ca = ca->next)
{
addNoiseTrack(noiseTrack, chunk, chunkSize, ca);
}
/* Step through base by base evaluating noise and reporting it if
* it's interesting. */
for (i=0; i<chunkSize; ++i)
{
struct noiseTrack *nt = &noiseTrack[i];
struct noise *noise = nt->noise;
int noiseCount = slCount(noise);
if (noiseCount > 1)
{
char commonVal;
int commonCount;
findCommon(noise, &commonVal, &commonCount);
if (commonCount*2 > noiseCount && commonVal != 'n')
{
double ratio = (double)commonCount/noiseCount;
double score;
ratio = ratio * ratio * ratio;
score = ratio * commonCount;
if (score >= 4.0)
{
fprintf(stdout, "%f %s:%d %c->%c in %d out of %d out of %d %s\n",
ratio*commonCount, chromName, i+baseOff+1,
chunk[i], commonVal,
commonCount, noiseCount, nt->cdnaCount, nt->noise->ca->cdna->srn->name);
fprintf(out, "%f %s:%d %c->%c in %d out of %d out of %d %s\n",
ratio*ratio*commonCount, chromName, i+baseOff+1,
chunk[i], commonVal,
commonCount, noiseCount, nt->cdnaCount, nt->noise->ca->cdna->srn->name);
}
}
}
}
freeCdnaAliList(&caList);
slFreeList(&cdnaNamesList);
freez(&dna);
recycleNoiseTrack(noiseTrack, chunkSize);
printf("%s %d maxNoise %d\n", chromName, baseOff, slCount(freeNoiseList));
}
}
return 0;
}
