int main(int argc, char **argv)
{
AjPAlign align;
AjPSeq a;
AjPSeq b;
AjPSeqout seqout;
AjPStr m;
AjPStr n;
AjPStr merged = NULL;
ajuint lena;
ajuint lenb;
const char *p;
const char *q;
ajint start1 = 0;
ajint start2 = 0;
float *path;
ajint *compass;
AjPMatrixf matrix;
AjPSeqCvt cvt = 0;
float **sub;
float gapopen;
float gapextend;
ajulong maxarr = 1000;
ajulong len; /* arbitrary. realloc'd if needed */
size_t stlen;
float score;
ajint begina;
ajint beginb;
embInit("merger", argc, argv);
a = ajAcdGetSeq("asequence");
b = ajAcdGetSeq("bsequence");
seqout = ajAcdGetSeqout("outseq");
matrix = ajAcdGetMatrixf("datafile");
gapopen = ajAcdGetFloat("gapopen");
gapextend = ajAcdGetFloat("gapextend");
align = ajAcdGetAlign("outfile");
gapopen = ajRoundFloat(gapopen, 8);
gapextend = ajRoundFloat(gapextend, 8);
AJCNEW(path, maxarr);
AJCNEW(compass, maxarr);
/*
** make the two sequences lowercase so we can show which one we are
** using in the merge by uppercasing it
*/
ajSeqFmtLower(a);
ajSeqFmtLower(b);
m = ajStrNew();
n = ajStrNew();
sub = ajMatrixfGetMatrix(matrix);
cvt = ajMatrixfGetCvt(matrix);
begina = ajSeqGetBegin(a);
beginb = ajSeqGetBegin(b);
lena = ajSeqGetLen(a);
lenb = ajSeqGetLen(b);
if(lenb > (ULONG_MAX/(ajulong)(lena+1)))
ajFatal("Sequences too big. Try 'supermatcher'");
len = lena*lenb;
if(len>maxarr)
{
ajDebug("merger: resize path, len to %d (%d * $d)\n",
len, lena, lenb);
stlen = (size_t) len;
AJCRESIZE(path,stlen);
AJCRESIZE(compass,stlen);
maxarr=len;
}
p = ajSeqGetSeqC(a);
q = ajSeqGetSeqC(b);
ajStrAssignC(&m,"");
ajStrAssignC(&n,"");
score = embAlignPathCalc(p,q,lena,lenb,gapopen,gapextend,path,sub,cvt,
compass, ajFalse);
/*score = embAlignScoreNWMatrix(path,compass,gapopen,gapextend,
a,b,lena,lenb,sub,cvt,
&start1,&start2);*/
embAlignWalkNWMatrix(path,a,b,&m,&n,lena,lenb, &start1,&start2,gapopen,
gapextend,compass);
/*
** now construct the merged sequence, uppercase the bits of the two
** input sequences which are used in the merger
*/
merger_Merge(align, &merged,p,q,m,n,start1,start2,
ajSeqGetNameC(a),ajSeqGetNameC(b));
embAlignReportGlobal(align, a, b, m, n,
start1, start2, gapopen, gapextend,
score, matrix, begina, beginb);
ajAlignWrite(align);
ajAlignReset(align);
/* write the merged sequence */
ajSeqAssignSeqS(a, merged);
ajSeqoutWriteSeq(seqout, a);
ajSeqoutClose(seqout);
ajSeqoutDel(&seqout);
ajSeqDel(&a);
ajSeqDel(&b);
ajAlignClose(align);
ajAlignDel(&align);
ajStrDel(&merged);
AJFREE(compass);
AJFREE(path);
ajStrDel(&n);
ajStrDel(&m);
embExit();
return 0;
}
int main(int argc, char **argv)
{
AjPAlign align;
AjPSeqall seqall;
AjPSeq a;
AjPSeq b;
AjPStr alga;
AjPStr algb;
AjPStr ss;
ajuint lena;
ajuint lenb;
const char *p;
const char *q;
ajint start1 = 0;
ajint start2 = 0;
ajint *compass;
float* ix;
float* iy;
float* m;
AjPMatrixf matrix;
AjPSeqCvt cvt = 0;
float **sub;
float gapopen;
float gapextend;
float endgapopen;
float endgapextend;
size_t maxarr = 1000; /* arbitrary. realloc'd if needed */
size_t len;
float score;
AjBool dobrief = ajTrue;
AjBool endweight = ajFalse; /* should end gap penalties be applied */
float id = 0.;
float sim = 0.;
float idx = 0.;
float simx = 0.;
AjPStr tmpstr = NULL;
embInit("needle", argc, argv);
matrix = ajAcdGetMatrixf("datafile");
a = ajAcdGetSeq("asequence");
ajSeqTrim(a);
seqall = ajAcdGetSeqall("bsequence");
gapopen = ajAcdGetFloat("gapopen");
gapextend = ajAcdGetFloat("gapextend");
endgapopen = ajAcdGetFloat("endopen");
endgapextend = ajAcdGetFloat("endextend");
dobrief = ajAcdGetBoolean("brief");
endweight = ajAcdGetBoolean("endweight");
align = ajAcdGetAlign("outfile");
gapopen = ajRoundFloat(gapopen, 8);
gapextend = ajRoundFloat(gapextend, 8);
AJCNEW0(compass, maxarr);
AJCNEW0(m, maxarr);
AJCNEW0(ix, maxarr);
AJCNEW0(iy, maxarr);
alga = ajStrNew();
algb = ajStrNew();
ss = ajStrNew();
sub = ajMatrixfGetMatrix(matrix);
cvt = ajMatrixfGetCvt(matrix);
lena = ajSeqGetLen(a);
while(ajSeqallNext(seqall,&b))
{
ajSeqTrim(b);
lenb = ajSeqGetLen(b);
if(lenb > (LONG_MAX/(size_t)(lena+1)))
ajDie("Sequences too big. Try 'stretcher'");
len = (size_t)lena*(size_t)lenb;
if(len>maxarr)
{
AJCRESIZETRY0(compass,(size_t)maxarr,len);
if(!compass)
ajDie("Sequences too big. Try 'stretcher'");
AJCRESIZETRY0(m,(size_t)maxarr,len);
if(!m)
ajDie("Sequences too big. Try 'stretcher'");
AJCRESIZETRY0(ix,(size_t)maxarr,len);
if(!ix)
ajDie("Sequences too big. Try 'stretcher'");
AJCRESIZETRY0(iy,(size_t)maxarr,len);
if(!iy)
ajDie("Sequences too big. Try 'stretcher'");
maxarr=len;
}
p = ajSeqGetSeqC(a);
q = ajSeqGetSeqC(b);
ajStrAssignC(&alga,"");
ajStrAssignC(&algb,"");
score = embAlignPathCalcWithEndGapPenalties(p, q, lena, lenb,
gapopen, gapextend, endgapopen, endgapextend,
&start1, &start2, sub, cvt,
m, ix, iy, compass, ajFalse, endweight);
embAlignWalkNWMatrixUsingCompass(p, q, &alga, &algb,
lena, lenb, &start1, &start2,
compass);
embAlignReportGlobal(align, a, b, alga, algb,
start1, start2,
gapopen, gapextend,
score, matrix,
ajSeqGetOffset(a), ajSeqGetOffset(b));
if(!dobrief)
{
embAlignCalcSimilarity(alga,algb,sub,cvt,lena,lenb,&id,&sim,&idx,
&simx);
ajFmtPrintS(&tmpstr,"Longest_Identity = %5.2f%%\n",
id);
ajFmtPrintAppS(&tmpstr,"Longest_Similarity = %5.2f%%\n",
sim);
ajFmtPrintAppS(&tmpstr,"Shortest_Identity = %5.2f%%\n",
idx);
ajFmtPrintAppS(&tmpstr,"Shortest_Similarity = %5.2f%%",
simx);
ajAlignSetSubHeaderApp(align, tmpstr);
}
ajAlignWrite(align);
ajAlignReset(align);
}
ajAlignClose(align);
ajAlignDel(&align);
ajSeqallDel(&seqall);
ajSeqDel(&a);
ajSeqDel(&b);
AJFREE(compass);
AJFREE(ix);
AJFREE(iy);
AJFREE(m);
ajStrDel(&alga);
ajStrDel(&algb);
ajStrDel(&ss);
ajStrDel(&tmpstr);
embExit();
return 0;
}
int main(int argc, char **argv)
{
/* ACD data item variables */
AjPFile hmmfile = NULL;
AjPSeqset seqfile = NULL;
AjPFile mapali = NULL;
AjPFile withali = NULL;
AjPAlign o = NULL;
AjBool m = ajFalse;
AjBool q = ajFalse;
/* Housekeeping variables */
AjPStr cmd = NULL;
AjPStr fmt = NULL;
AjBool fmtok = ajFalse;
AjPStr rnd = NULL;
AjPSeqout rndo = NULL;
/* ACD file processing */
embInitPV("ehmmalign",argc,argv,"HMMERNEW",VERSION);
hmmfile = ajAcdGetInfile("hmmfile");
seqfile = ajAcdGetSeqset("seqfile");
mapali = ajAcdGetInfile("mapali");
withali = ajAcdGetInfile("withali");
o = ajAcdGetAlign("o");
m = ajAcdGetBoolean("m");
q = ajAcdGetBoolean("q");
/* MAIN APPLICATION CODE */
/* 1. Housekeeping */
cmd = ajStrNew();
fmt = ajStrNew();
rnd = ajStrNew();
/* 2. Re-write seqfile to a temporary file in a format (fasta) HMMER can understand.
We cannot just pass the name of seqfile to HMMER as the name provided might be a
USA which HMMER would not understand. */
ajFilenameSetTempname(&rnd);
rndo = ajSeqoutNew();
if(!ajSeqoutOpenFilename(rndo, rnd))
ajFatal("Terminal ajSeqFileNewOut failure. Email EMBOSS helpdesk!\n");
ajSeqoutSetFormatC(rndo, "fasta");
ajSeqoutWriteSet(rndo, seqfile);
ajSeqoutClose(rndo);
ajSeqoutDel(&rndo);
/* 3. Build hmmalign command line */
/* Command line is built in this order:
i. Application name.
ii. HMMER 'options' (in order they appear in ACD file)
iii.HMMER 'options' (that don't appear in ACD file)
iv. HMMER & new parameters.
*/
ajFmtPrintS(&cmd, "%S ", ajAcdGetpathC("hmmalign"));
if(mapali)
ajFmtPrintAppS(&cmd, " --mapali %s ", ajFileGetNameC(mapali));
if(withali)
ajFmtPrintAppS(&cmd, " --withali %s ", ajFileGetNameC(withali));
if(m)
ajStrAppendC(&cmd, " -m ");
if(q)
ajStrAppendC(&cmd, " -q ");
/* Ensure output alignment is in user-specified format. */
fmtok=ajTrue;
ajStrAssignS(&fmt, ajAlignGetFormat(o));
/* fasta and a2m are identical formats. */
if(ajStrMatchC(fmt, "fasta"))
ajStrAssignC(&fmt, "A2M");
else if(ajStrMatchC(fmt, "a2m"))
ajStrAssignC(&fmt, "A2M");
else if(ajStrMatchC(fmt, "msf"))
ajStrAssignC(&fmt, "MSF");
else if(ajStrMatchC(fmt, "phylip"))
ajStrAssignC(&fmt, "PHYLIP");
/* hmmer also supports stockholm, SELEX & Clustal output, EMBOSS does not.
EMBOSS supports unknown/multiple/simple and srs output, hmmer does not. */
else
fmtok = ajFalse;
if(!fmtok)
{
/* This could be replaced with code to reformat the file. */
ajWarn("Specified output alignment format ('o' ACD option) is "
"not understood by HMMER. Using stockholm format instead.");
ajStrAssignC(&fmt, "Stockholm");
}
/* rnd is the name of the rewritten seqfile. MUST specify FASTA format explicitly. */
ajFmtPrintAppS(&cmd, " --informat FASTA --outformat %S -o %s %s %S",
fmt,
ajAlignGetFilename(o),
ajFileGetNameC(hmmfile),
rnd);
/* 4. Close ACD files */
ajFileClose(&hmmfile);
ajSeqsetDel(&seqfile);
ajFileClose(&mapali);
ajFileClose(&withali);
ajAlignClose(o);
ajAlignDel(&o);
/* 5. Call hmmalign */
ajFmtPrint("\n%S\n\n", cmd);
ajSysExecS(cmd);
/* 6. Exit cleanly */
ajSysFileUnlinkS(rnd);
ajStrDel(&cmd);
ajStrDel(&fmt);
ajStrDel(&rnd);
embExit();
return 0;
}
int main(int argc, char **argv)
{
AjPSeqall queryseqs;
AjPSeqset targetseqs;
AjPSeq queryseq;
const AjPSeq targetseq;
AjPStr queryaln = 0;
AjPStr targetaln = 0;
AjPFile errorf;
AjBool show = ajFalse;
const char *queryseqc;
const char *targetseqc;
AjPMatrixf matrix;
AjPSeqCvt cvt = 0;
float **sub;
ajint *compass = NULL;
float *path = NULL;
float gapopen;
float gapextend;
float score;
float minscore;
ajuint j, k;
ajint querystart = 0;
ajint targetstart = 0;
ajint queryend = 0;
ajint targetend = 0;
ajint width = 0;
AjPTable kmers = 0;
ajint wordlen = 6;
ajint oldmax = 0;
ajint newmax = 0;
ajuint ntargetseqs;
ajuint nkmers;
AjPAlign align = NULL;
EmbPWordMatch maxmatch; /* match with maximum score */
/* Cursors for the current sequence being scanned,
** i.e., until which location it was scanned.
** Separate cursor/location entries for each sequence in the seqset.
*/
ajuint* lastlocation;
EmbPWordRK* wordsw = NULL;
AjPList* matchlist = NULL;
embInit("supermatcher", argc, argv);
matrix = ajAcdGetMatrixf("datafile");
queryseqs = ajAcdGetSeqall("asequence");
targetseqs= ajAcdGetSeqset("bsequence");
gapopen = ajAcdGetFloat("gapopen");
gapextend = ajAcdGetFloat("gapextend");
wordlen = ajAcdGetInt("wordlen");
align = ajAcdGetAlign("outfile");
errorf = ajAcdGetOutfile("errorfile");
width = ajAcdGetInt("width"); /* width for banded Smith-Waterman */
minscore = ajAcdGetFloat("minscore");
gapopen = ajRoundFloat(gapopen, 8);
gapextend = ajRoundFloat(gapextend, 8);
sub = ajMatrixfGetMatrix(matrix);
cvt = ajMatrixfGetCvt(matrix);
embWordLength(wordlen);
/* seqset sequence is the reference sequence for SAM format */
ajAlignSetRefSeqIndx(align, 1);
ajSeqsetTrim(targetseqs);
ntargetseqs = ajSeqsetGetSize(targetseqs);
AJCNEW0(matchlist, ntargetseqs);
/* get tables of words */
for(k=0;k<ntargetseqs;k++)
{
targetseq = ajSeqsetGetseqSeq(targetseqs, k);
embWordGetTable(&kmers, targetseq);
ajDebug("Number of distinct kmers found so far: %d\n",
ajTableGetLength(kmers));
}
AJCNEW0(lastlocation, ntargetseqs);
if(ajTableGetLength(kmers)<1)
ajErr("no kmers found");
nkmers = embWordRabinKarpInit(kmers, &wordsw, wordlen, targetseqs);
while(ajSeqallNext(queryseqs,&queryseq))
{
ajSeqTrim(queryseq);
queryaln = ajStrNewRes(1+ajSeqGetLen(queryseq));
ajDebug("Read '%S'\n", ajSeqGetNameS(queryseq));
for(k=0;k<ntargetseqs;k++)
{
lastlocation[k]=0;
matchlist[k] = ajListstrNew();
}
embWordRabinKarpSearch(ajSeqGetSeqS(queryseq), targetseqs,
(const EmbPWordRK*)wordsw, wordlen, nkmers,
matchlist, lastlocation, ajFalse);
for(k=0;k<ajSeqsetGetSize(targetseqs);k++)
{
targetseq = ajSeqsetGetseqSeq(targetseqs, k);
ajDebug("Processing '%S'\n", ajSeqGetNameS(targetseq));
if(ajListGetLength(matchlist[k])==0)
{
ajFmtPrintF(errorf,
"No wordmatch start points for "
"%s vs %s. No alignment\n",
ajSeqGetNameC(queryseq),ajSeqGetNameC(targetseq));
embWordMatchListDelete(&matchlist[k]);
continue;
}
/* only the maximum match is used as seed
* (if there is more than one location with the maximum match
* only the first one is used)
* TODO: we should add a new option to make above limit optional
*/
maxmatch = embWordMatchFirstMax(matchlist[k]);
supermatcher_findendpoints(maxmatch,targetseq, queryseq,
&targetstart, &querystart,
&targetend, &queryend);
targetaln=ajStrNewRes(1+ajSeqGetLen(targetseq));
queryseqc = ajSeqGetSeqC(queryseq);
targetseqc = ajSeqGetSeqC(targetseq);
ajStrAssignC(&queryaln,"");
ajStrAssignC(&targetaln,"");
ajDebug("++ %S v %S start:%d %d end:%d %d\n",
ajSeqGetNameS(targetseq), ajSeqGetNameS(queryseq),
targetstart, querystart, targetend, queryend);
newmax = (targetend-targetstart+2)*width;
if(newmax > oldmax)
{
AJCRESIZE0(path,oldmax,newmax);
AJCRESIZE0(compass,oldmax,newmax);
oldmax=newmax;
ajDebug("++ memory re/allocation for path/compass arrays"
" to size: %d\n", newmax);
}
else
{
AJCSET0(path,newmax);
AJCSET0(compass,newmax);
}
ajDebug("Calling embAlignPathCalcSWFast "
"%d..%d [%d/%d] %d..%d [%d/%d] width:%d\n",
querystart, queryend, (queryend - querystart + 1),
ajSeqGetLen(queryseq),
targetstart, targetend, (targetend - targetstart + 1),
ajSeqGetLen(targetseq),
width);
score = embAlignPathCalcSWFast(&targetseqc[targetstart],
&queryseqc[querystart],
targetend-targetstart+1,
queryend-querystart+1,
0,width,
gapopen,gapextend,
path,sub,cvt,
compass,show);
if(score>minscore)
{
embAlignWalkSWMatrixFast(path,compass,gapopen,gapextend,
targetseq,queryseq,
&targetaln,&queryaln,
targetend-targetstart+1,
queryend-querystart+1,
0,width,
&targetstart,&querystart);
if(!ajAlignFormatShowsSequences(align))
{
ajAlignDefineCC(align, ajStrGetPtr(targetaln),
ajStrGetPtr(queryaln),
ajSeqGetNameC(targetseq),
ajSeqGetNameC(queryseq));
ajAlignSetScoreR(align, score);
}
else
{
ajDebug(" queryaln:%S \ntargetaln:%S\n",
queryaln,targetaln);
embAlignReportLocal(align,
queryseq, targetseq,
queryaln, targetaln,
querystart, targetstart,
gapopen, gapextend,
score, matrix,
1 + ajSeqGetOffset(queryseq),
1 + ajSeqGetOffset(targetseq)
);
}
ajAlignWrite(align);
ajAlignReset(align);
}
ajStrDel(&targetaln);
embWordMatchListDelete(&matchlist[k]);
}
ajStrDel(&queryaln);
}
for(k=0;k<nkmers;k++)
{
AJFREE(wordsw[k]->seqindxs);
AJFREE(wordsw[k]->nSeqMatches);
for(j=0;j<wordsw[k]->nseqs;j++)
AJFREE(wordsw[k]->locs[j]);
AJFREE(wordsw[k]->nnseqlocs);
AJFREE(wordsw[k]->locs);
AJFREE(wordsw[k]);
}
embWordFreeTable(&kmers);
if(!ajAlignFormatShowsSequences(align))
ajMatrixfDel(&matrix);
AJFREE(path);
AJFREE(compass);
AJFREE(kmers);
AJFREE(wordsw);
AJFREE(matchlist);
AJFREE(lastlocation);
ajAlignClose(align);
ajAlignDel(&align);
ajSeqallDel(&queryseqs);
ajSeqDel(&queryseq);
ajSeqsetDel(&targetseqs);
ajFileClose(&errorf);
embExit();
return 0;
}
int main(int argc, char **argv)
{
AjPSeqset seqset;
AjPSeqall seqall;
AjPSeq queryseq;
const AjPSeq targetseq;
ajint wordlen;
AjPTable wordsTable = NULL;
AjPList* matchlist = NULL;
AjPFile logfile;
AjPFeattable* seqsetftables = NULL;
AjPFeattable seqallseqftable = NULL;
AjPFeattabOut ftoutforseqsetseq = NULL;
AjPFeattabOut ftoutforseqallseq = NULL;
AjPAlign align = NULL;
AjIList iter = NULL;
ajint targetstart;
ajint querystart;
ajint len;
ajuint i, j;
ajulong nAllMatches = 0;
ajulong sumAllScore = 0;
AjBool dumpAlign = ajTrue;
AjBool dumpFeature = ajTrue;
AjBool checkmode = ajFalse;
EmbPWordRK* wordsw = NULL;
ajuint npatterns = 0;
ajuint seqsetsize;
ajuint nmatches;
ajuint* nmatchesseqset;
ajuint* lastlocation; /* Cursors for Rabin-Karp search. */
/* Shows until what point the query sequence was
* scanned for a pattern sequences in the seqset.
*/
char* paddedheader = NULL;
const char* header;
AjPStr padding;
header = "Pattern %S #pat-sequences #all-matches avg-match-length\n";
padding = ajStrNew();
embInit("wordmatch", argc, argv);
wordlen = ajAcdGetInt("wordsize");
seqset = ajAcdGetSeqset("asequence");
seqall = ajAcdGetSeqall("bsequence");
logfile = ajAcdGetOutfile("logfile");
dumpAlign = ajAcdGetToggle("dumpalign");
dumpFeature = ajAcdGetToggle("dumpfeat");
if(dumpAlign)
{
align = ajAcdGetAlign("outfile");
ajAlignSetExternal(align, ajTrue);
}
seqsetsize = ajSeqsetGetSize(seqset);
ajSeqsetTrim(seqset);
AJCNEW0(matchlist, seqsetsize);
AJCNEW0(seqsetftables, seqsetsize);
AJCNEW0(nmatchesseqset, seqsetsize);
if (dumpFeature)
{
ftoutforseqsetseq = ajAcdGetFeatout("aoutfeat");
ftoutforseqallseq = ajAcdGetFeatout("boutfeat");
}
checkmode = !dumpFeature && !dumpAlign;
embWordLength(wordlen);
ajFmtPrintF(logfile, "Small sequence/file for constructing"
" target patterns: %S\n", ajSeqsetGetUsa(seqset));
ajFmtPrintF(logfile, "Large sequence/file to be scanned"
" for patterns: %S\n", ajSeqallGetUsa(seqall));
ajFmtPrintF(logfile, "Number of sequences in the patterns file: %u\n",
seqsetsize);
ajFmtPrintF(logfile, "Pattern/word length: %u\n", wordlen);
for(i=0;i<seqsetsize;i++)
{
targetseq = ajSeqsetGetseqSeq(seqset, i);
embWordGetTable(&wordsTable, targetseq);
}
AJCNEW0(lastlocation, seqsetsize);
if(ajTableGetLength(wordsTable)>0)
{
npatterns = embWordRabinKarpInit(wordsTable,
&wordsw, wordlen, seqset);
ajFmtPrintF(logfile, "Number of patterns/words found: %u\n", npatterns);
while(ajSeqallNext(seqall,&queryseq))
{
for(i=0;i<seqsetsize;i++)
{
lastlocation[i]=0;
if (!checkmode)
matchlist[i] = ajListstrNew();
}
nmatches = embWordRabinKarpSearch(
ajSeqGetSeqS(queryseq), seqset,
(EmbPWordRK const *)wordsw, wordlen, npatterns,
matchlist, lastlocation, checkmode);
nAllMatches += nmatches;
if (checkmode)
continue;
for(i=0;i<seqsetsize;i++)
{
if(ajListGetLength(matchlist[i])>0)
{
iter = ajListIterNewread(matchlist[i]) ;
while(embWordMatchIter(iter, &targetstart, &querystart, &len,
&targetseq))
{
if(dumpAlign)
{
ajAlignDefineSS(align, targetseq, queryseq);
ajAlignSetScoreI(align, len);
/* ungapped alignment means same length
* for both sequences
*/
ajAlignSetSubRange(align, targetstart, 1, len,
ajSeqIsReversed(targetseq),
ajSeqGetLen(targetseq),
querystart, 1, len,
ajSeqIsReversed(queryseq),
ajSeqGetLen(queryseq));
}
}
if(dumpAlign)
{
ajAlignWrite(align);
ajAlignReset(align);
}
if(ajListGetLength(matchlist[i])>0 && dumpFeature)
{
embWordMatchListConvToFeat(matchlist[i],
&seqsetftables[i],
&seqallseqftable,
targetseq, queryseq);
ajFeattableWrite(ftoutforseqallseq, seqallseqftable);
ajFeattableDel(&seqallseqftable);
}
ajListIterDel(&iter);
}
embWordMatchListDelete(&matchlist[i]);
}
}
/* search completed, now report statistics */
for(i=0;i<npatterns;i++)
{
sumAllScore += wordsw[i]->lenMatches;
for(j=0;j<wordsw[i]->nseqs;j++)
nmatchesseqset[wordsw[i]->seqindxs[j]] +=
wordsw[i]->nSeqMatches[j];
}
ajFmtPrintF(logfile, "Number of sequences in the file scanned "
"for patterns: %u\n", ajSeqallGetCount(seqall));
ajFmtPrintF(logfile, "Number of all matches: %Lu"
" (wordmatch finds exact matches only)\n", nAllMatches);
if(nAllMatches>0)
{
ajFmtPrintF(logfile, "Sum of match lengths: %Lu\n", sumAllScore);
ajFmtPrintF(logfile, "Average match length: %.2f\n",
sumAllScore*1.0/nAllMatches);
ajFmtPrintF(logfile, "\nDistribution of the matches among pattern"
" sequences:\n");
ajFmtPrintF(logfile, "-----------------------------------------"
"-----------\n");
for(i=0;i<ajSeqsetGetSize(seqset);i++)
{
if (nmatchesseqset[i]>0)
ajFmtPrintF(logfile, "%-42s: %8u\n",
ajSeqGetNameC(ajSeqsetGetseqSeq(seqset, i)),
nmatchesseqset[i]);
ajFeattableWrite(ftoutforseqsetseq, seqsetftables[i]);
ajFeattableDel(&seqsetftables[i]);
}
ajFmtPrintF(logfile, "\nPattern statistics:\n");
ajFmtPrintF(logfile, "-------------------\n");
if(wordlen>7)
ajStrAppendCountK(&padding, ' ', wordlen-7);
paddedheader = ajFmtString(header,padding);
ajFmtPrintF(logfile, paddedheader);
for(i=0;i<npatterns;i++)
if (wordsw[i]->nMatches>0)
ajFmtPrintF(logfile, "%-7s: %12u %12u %17.2f\n",
wordsw[i]->word->fword, wordsw[i]->nseqs,
wordsw[i]->nMatches,
wordsw[i]->lenMatches*1.0/wordsw[i]->nMatches);
}
}
for(i=0;i<npatterns;i++)
{
for(j=0;j<wordsw[i]->nseqs;j++)
AJFREE(wordsw[i]->locs[j]);
AJFREE(wordsw[i]->locs);
AJFREE(wordsw[i]->seqindxs);
AJFREE(wordsw[i]->nnseqlocs);
AJFREE(wordsw[i]->nSeqMatches);
AJFREE(wordsw[i]);
}
embWordFreeTable(&wordsTable);
AJFREE(wordsw);
AJFREE(matchlist);
AJFREE(lastlocation);
AJFREE(nmatchesseqset);
AJFREE(seqsetftables);
if(dumpAlign)
{
ajAlignClose(align);
ajAlignDel(&align);
}
if(dumpFeature)
{
ajFeattabOutDel(&ftoutforseqsetseq);
ajFeattabOutDel(&ftoutforseqallseq);
}
ajFileClose(&logfile);
ajSeqallDel(&seqall);
ajSeqsetDel(&seqset);
ajSeqDel(&queryseq);
ajStrDel(&padding);
AJFREE(paddedheader);
embExit();
return 0;
}
int main(int argc, char **argv)
{
AjPSeqall seqall;
AjPSeq seq = NULL;
AjPFile inf = NULL;
AjPStr strand = NULL;
AjPStr substr = NULL;
AjPStr name = NULL;
AjPStr mname = NULL;
AjPStr tname = NULL;
AjPStr pname = NULL;
AjPStr line = NULL;
AjPStr cons = NULL;
AjPStr m = NULL;
AjPStr n = NULL;
AjPAlign align= NULL; /* JISON, replaces AjPOutfile outf */
ajint type;
ajint begin;
ajint end;
ajulong len;
ajint i;
ajint j;
float **fmatrix=NULL;
ajint mlen;
float maxfs;
ajint thresh;
float gapopen;
float gapextend;
float opencoeff;
float extendcoeff;
const char *p;
ajulong maxarr = 1000;
ajulong alen;
float *path;
ajint *compass;
size_t stlen;
embInit("prophet", argc, argv);
seqall = ajAcdGetSeqall("sequence");
inf = ajAcdGetInfile("infile");
opencoeff = ajAcdGetFloat("gapopen");
extendcoeff = ajAcdGetFloat("gapextend");
align = ajAcdGetAlign("outfile"); /*JISON replacing outfile */
opencoeff = ajRoundFloat(opencoeff, 8);
extendcoeff = ajRoundFloat(extendcoeff, 8);
substr = ajStrNew();
name = ajStrNew();
mname = ajStrNew();
tname = ajStrNew();
line = ajStrNew();
m = ajStrNewC("");
n = ajStrNewC("");
type = prophet_getType(inf,&tname);
if(!type)
ajFatal("Unrecognised profile/matrix file format");
prophet_read_profile(inf,&pname,&mname,&mlen,&gapopen,&gapextend,&thresh,
&maxfs, &cons);
ajAlignSetMatrixName(align, mname);
AJCNEW(fmatrix, mlen);
for(i=0;i<mlen;++i)
{
AJCNEW(fmatrix[i], AZ);
if(!ajReadlineTrim(inf,&line))
ajFatal("Missing matrix line");
p = ajStrGetPtr(line);
p = ajSysFuncStrtok(p," \t");
for(j=0;j<AZ;++j)
{
sscanf(p,"%f",&fmatrix[i][j]);
p = ajSysFuncStrtok(NULL," \t");
}
}
AJCNEW(path, maxarr);
AJCNEW(compass, maxarr);
while(ajSeqallNext(seqall, &seq))
{
begin = ajSeqallGetseqBegin(seqall);
end = ajSeqallGetseqEnd(seqall);
ajStrAssignC(&name,ajSeqGetNameC(seq));
strand = ajSeqGetSeqCopyS(seq);
ajStrAssignSubC(&substr,ajStrGetPtr(strand),begin-1,end-1);
len = ajStrGetLen(substr);
if(len > (ULONG_MAX/(ajulong)(mlen+1)))
ajFatal("Sequences too big. Try 'supermatcher'");
alen = len*mlen;
if(alen>maxarr)
{
stlen = (size_t) alen;
AJCRESIZE(path,stlen);
AJCRESIZE(compass,stlen);
maxarr=alen;
}
ajStrAssignC(&m,"");
ajStrAssignC(&n,"");
/* JISON used to be
prophet_scan_profile(substr,pname,name,mlen,fmatrix,
outf,cons,opencoeff,
extendcoeff,path,compass,&m,&n,len); */
/* JISON new call and reset align */
prophet_scan_profile(substr,name,pname,mlen,fmatrix,
align,cons,opencoeff,
extendcoeff,path,compass,&m,&n,(ajint)len);
ajAlignReset(align);
ajStrDel(&strand);
}
for(i=0;i<mlen;++i)
AJFREE (fmatrix[i]);
AJFREE (fmatrix);
AJFREE(path);
AJFREE(compass);
ajStrDel(&line);
ajStrDel(&cons);
ajStrDel(&name);
ajStrDel(&pname);
ajStrDel(&mname);
ajStrDel(&tname);
ajStrDel(&substr);
ajStrDel(&m);
ajStrDel(&n);
ajSeqDel(&seq);
ajFileClose(&inf);
ajAlignClose(align);
ajAlignDel(&align);
ajSeqallDel(&seqall);
embExit();
return 0;
}