int main(int argc, char **argv)
{
/* Variable Declarations */
AjPSeqset seqset = NULL;
AjPMatrixf fmat = NULL;
float thresh;
float threshlow;
float threshup;
float gapopen;
float gapextend;
AjPSeqout seqout = NULL;
AjPSeqout seqoutred = NULL;
AjPStr mode = NULL;
ajint moden;
ajuint i;
/* toggle "feature" from ACD not retrieved ... no need */
const AjPSeq seq = NULL;
AjPList list = NULL; /* List for redundancy removal. */
AjPUint keep = NULL; /* 1: Sequence in list was non-redundant,
0: redundant. */
ajuint nseq = 0; /* No. seqs. in list. */
ajint nseqnr = 0; /* No. non-redundant seqs. in list. */
/* ACD File Processing */
embInit("skipredundant", argc, argv);
seqset = ajAcdGetSeqset("sequences");
mode = ajAcdGetListSingle("mode");
fmat = ajAcdGetMatrixf("datafile");
thresh = ajAcdGetFloat("threshold");
threshlow = ajAcdGetFloat("minthreshold");
threshup = ajAcdGetFloat("maxthreshold");
gapopen = ajAcdGetFloat("gapopen");
gapextend = ajAcdGetFloat("gapextend");
seqout = ajAcdGetSeqoutall("outseq");
seqoutred = ajAcdGetSeqoutall("redundantoutseq");
/* Application logic */
list = ajListNew();
skipredundant_SeqsetToList(list, seqset);
keep = ajUintNew();
ajStrToInt(mode, &moden);
if(moden == 1)
/* Remove redundancy at a single threshold % sequence similarity */
{
if((!embDmxSeqNR(list, &keep, &nseqnr, fmat, gapopen,
gapextend, thresh, ajFalse)))
ajFatal("embDmxSeqNR unexpected failure!");
}
else if (moden == 2)
/* 2: Remove redundancy outside a range of acceptable threshold % similarity */
{
if((!embDmxSeqNRRange(list, &keep, &nseqnr, fmat, gapopen,
gapextend, threshlow, threshup, ajFalse)))
ajFatal("embDmxSeqNRRange unexpected failure!");
}
else
ajFatal("Invalid mode (not 1 or 2) which should never occur (check ACD file!)");
nseq = ajSeqsetGetSize(seqset);
for(i=0; i<nseq; i++)
{
seq = ajSeqsetGetseqSeq(seqset, i);
if(ajUintGet(keep, i))
ajSeqoutWriteSeq(seqout, seq);
else if(seqoutred)
ajSeqoutWriteSeq(seqoutred, seq);
}
/* Memory management and exit */
ajSeqsetDel(&seqset);
ajMatrixfDel(&fmat);
ajStrDel(&mode);
ajSeqoutClose(seqout);
ajSeqoutDel(&seqout);
if(seqoutred)
{
ajSeqoutClose(seqoutred);
ajSeqoutDel(&seqoutred);
}
skipredundant_ClearList(list);
ajListFree(&list);
ajUintDel(&keep);
embExit();
return 0;
}
int main(int argc, char **argv)
{
AjPList sigin = NULL; /* Signature input file names. */
AjPStr signame = NULL; /* Name of signature file. */
AjPFile sigf = NULL; /* Signature input file. */
EmbPSignature sig = NULL; /* Signature. */
AjPList siglist = NULL; /* List of signatures. */
AjIList sigiter = NULL; /* Iterator for siglist. */
AjBool sigok = ajFalse; /* True if signature processed ok. */
EmbPHit hit = NULL; /* Hit to store signature-sequence match. */
AjPList hits = NULL; /* List of hits */
AjPList ligands = NULL; /* List of top-scoring ligands. */
AjPSeqall database=NULL; /* Protein sequences to match signature
against. */
AjPSeq seq = NULL; /* Current sequence. */
AjPMatrixf sub =NULL; /* Residue substitution matrix. */
float gapo =0.0; /* Gap insertion penalty. */
float gape =0.0; /* Gap extension penalty. */
AjPStr nterm=NULL; /* Holds N-terminal matching options from
acd. */
ajint ntermi=0; /* N-terminal option as int. */
AjPFile hitsf =NULL; /* Hits output file.
sequence matches. */
AjPDirout hitsdir=NULL; /* Directory of hits files (output). */
AjPFile alignf =NULL; /* Alignment output file. */
AjPDirout aligndir=NULL; /* Directory of alignment files (output). */
AjPFile resultsf =NULL; /* Results file (output). */
AjPDirout resultsdir=NULL; /* Directory of results files (output). */
AjPStr mode = NULL; /* Mode, 1: Patch score mode, 2:
Site score mode. */
ajint modei = 0; /* Selected mode as integer. */
SigPLighit lighit = NULL;
embInitPV("sigscanlig", argc, argv, "SIGNATURE",VERSION);
/* GET VALUES FROM ACD */
sigin = ajAcdGetDirlist("siginfilesdir");
database = ajAcdGetSeqall("dbseqall");
sub = ajAcdGetMatrixf("sub");
gapo = ajAcdGetFloat("gapo");
gape = ajAcdGetFloat("gape");
nterm = ajAcdGetListSingle("nterm");
hitsdir = ajAcdGetOutdir("hitsoutdir");
aligndir = ajAcdGetOutdir("alignoutdir");
resultsdir = ajAcdGetOutdir("resultsoutdir");
mode = ajAcdGetListSingle("mode");
/*Assign N-terminal matching option etc. */
ajFmtScanS(nterm, "%d", &ntermi);
modei = (ajint) ajStrGetCharFirst(mode)-48;
/* READ & PROCESS SIGNATURES */
siglist = ajListNew();
while(ajListPop(sigin, (void **) &signame))
{
/* Read signature files, compile signatures and populate list. */
sigok = ajFalse;
if((sigf = ajFileNewInNameS(signame)))
if((sig = embSignatureReadNew(sigf)))
if(embSignatureCompile(&sig, gapo, gape, sub))
{
sigok=ajTrue;
ajListPushAppend(siglist, sig);
/*
ajFmtPrint("Id: %S\nDomid: %S\nLigid: %S\nns: %d\n"
"sn: %d\nnp: %d\npn: %d\nminpatch: %d\n"
"maxgap: %d\n",
sig->Id, sig->Domid, sig->Ligid, sig->ns,
sig->sn, sig->np, sig->pn, sig->minpatch,
sig->maxgap); */
}
if(!sigok)
{
ajWarn("Could not process %S", signame);
embSignatureDel(&sig);
ajFileClose(&sigf);
ajStrDel(&signame);
continue;
}
ajFileClose(&sigf);
ajStrDel(&signame);
}
ajListFree(&sigin);
/* ALIGN EACH QUERY SEQUENCE TO LIST OF SIGNATURE */
while(ajSeqallNext(database, &seq))
{
/* Do sequence-signature alignment and save results */
hits = ajListNew();
sigiter = ajListIterNew(siglist);
while((sig = (EmbPSignature) ajListIterGet(sigiter)))
{
if(embSignatureAlignSeq(sig, seq, &hit, ntermi))
{
hit->Sig = sig;
ajListPushAppend(hits, hit);
hit=NULL; /* To force reallocation by embSignatureAlignSeq */
}
/* There has to be a hit for each signature for correct
generation of the LHF by sigscanlig_WriteFasta. So push
an empty hit if necessary. 'hit'=NULL forces
reallocation by embSignatureAlignSeq. */
/*
else
{
hit = embHitNew();
ajListPushAppend(hits, hit);
hit=NULL;
}
*/
}
ajListIterDel(&sigiter);
/* Rank-order the list of hits by score */
ajListSort(hits, embMatchinvScore);
/* Write ligand hits & alignment files (output) */
hitsf = ajFileNewOutNameDirS(ajSeqGetNameS(seq), hitsdir);
alignf = ajFileNewOutNameDirS(ajSeqGetNameS(seq), aligndir);
resultsf = ajFileNewOutNameDirS(ajSeqGetNameS(seq), resultsdir);
/* if((!sigscanlig_WriteFasta(hitsf, siglist, hits)))
ajFatal("Bad args to sigscanlig_WriteFasta"); */
if((!sigscanlig_WriteFasta(hitsf, hits)))
ajFatal("Bad args to sigscanlig_WriteFasta");
if((!sigscanlig_SignatureAlignWriteBlock(alignf, hits)))
ajFatal("Bad args to sigscanlig_SignatureAlignWriteBlock");
/* if((!sigscanlig_SignatureAlignWriteBlock(alignf, siglist, hits)))
ajFatal("Bad args to sigscanlig_SignatureAlignWriteBlock"); */
/* Sort list of hits by ligand type and site number.
Process list of ligands and print out. */
ajListSortTwo(hits, embMatchLigid, embMatchSN);
if(modei==1)
ligands = sigscanlig_score_ligands_patch(hits);
else if(modei==2)
ligands = sigscanlig_score_ligands_site(hits);
else
ajFatal("Unrecognised mode");
sigscanlig_WriteResults(ligands, resultsf);
ajFileClose(&hitsf);
ajFileClose(&alignf);
ajFileClose(&resultsf);
/* Memory management */
while(ajListPop(hits, (void **) &hit))
embHitDel(&hit);
ajListFree(&hits);
while(ajListPop(ligands, (void **) &lighit))
sigscanlig_LigHitDel(&lighit);
ajListFree(&ligands);
}
/* MEMORY MANAGEMENT */
while(ajListPop(siglist, (void **) &sig))
embSignatureDel(&sig);
ajListFree(&siglist);
ajSeqallDel(&database);
ajMatrixfDel(&sub);
ajStrDel(&nterm);
ajDiroutDel(&hitsdir);
ajDiroutDel(&aligndir);
ajDiroutDel(&resultsdir);
ajStrDel(&mode);
embExit();
return 0;
}
int main(int argc, char **argv)
{
AjPSeqall seq1;
AjPSeqset seq2;
AjPSeq a;
const AjPSeq b;
AjPStr m = 0;
AjPStr n = 0;
AjPFile errorf;
AjBool show = ajFalse;
ajint lena = 0;
ajint lenb = 0;
const char *p;
const char *q;
AjPMatrixf matrix;
AjPSeqCvt cvt = 0;
float **sub;
ajint *compass = NULL;
float *path = NULL;
float gapopen;
float gapextend;
float score;
ajint begina;
ajint i;
ajuint k;
ajint beginb;
ajint start1 = 0;
ajint start2 = 0;
ajint end1 = 0;
ajint end2 = 0;
ajint width = 0;
AjPTable seq1MatchTable = 0;
ajint wordlen = 6;
ajint oldmax = 0;
AjPAlign align = NULL;
embInit("supermatcher", argc, argv);
matrix = ajAcdGetMatrixf("datafile");
seq1 = ajAcdGetSeqall("asequence");
seq2 = ajAcdGetSeqset("bsequence");
gapopen = ajAcdGetFloat("gapopen");
gapextend = ajAcdGetFloat("gapextend");
wordlen = ajAcdGetInt("wordlen");
align = ajAcdGetAlign("outfile");
errorf = ajAcdGetOutfile("errorfile");
width = ajAcdGetInt("width"); /* not the same as awidth */
gapopen = ajRoundFloat(gapopen, 8);
gapextend = ajRoundFloat(gapextend, 8);
sub = ajMatrixfGetMatrix(matrix);
cvt = ajMatrixfGetCvt(matrix);
embWordLength(wordlen);
ajSeqsetTrim(seq2);
while(ajSeqallNext(seq1,&a))
{
ajSeqTrim(a);
begina = 1 + ajSeqGetOffset(a);
m = ajStrNewRes(1+ajSeqGetLen(a));
lena = ajSeqGetLen(a);
ajDebug("Read '%S'\n", ajSeqGetNameS(a));
if(!embWordGetTable(&seq1MatchTable, a)) /* get table of words */
ajErr("Could not generate table for %s\n",
ajSeqGetNameC(a));
for(k=0;k<ajSeqsetGetSize(seq2);k++)
{
b = ajSeqsetGetseqSeq(seq2, k);
lenb = ajSeqGetLen(b);
beginb = 1 + ajSeqGetOffset(b);
ajDebug("Processing '%S'\n", ajSeqGetNameS(b));
p = ajSeqGetSeqC(a);
q = ajSeqGetSeqC(b);
if(!supermatcher_findstartpoints(seq1MatchTable,b,a,
&start1, &start2,
&end1, &end2))
{
ajFmtPrintF(errorf,
"No wordmatch start points for "
"%s vs %s. No alignment\n",
ajSeqGetNameC(a),ajSeqGetNameC(b));
continue;
}
n=ajStrNewRes(1+ajSeqGetLen(b));
ajStrAssignC(&m,"");
ajStrAssignC(&n,"");
ajDebug("++ %S v %S start:%d %d end:%d %d\n",
ajSeqGetNameS(a), ajSeqGetNameS(b),
start1, start2, end1, end2);
if(end1-start1+1 > oldmax)
{
oldmax = ((end1-start1)+1);
AJRESIZE(path,oldmax*width*sizeof(float));
AJRESIZE(compass,oldmax*width*sizeof(ajint));
ajDebug("++ resize to oldmax: %d\n", oldmax);
}
for(i=0;i<((end1-start1)+1)*width;i++)
path[i] = 0.0;
ajDebug("Calling embAlignPathCalcFast "
"%d..%d [%d/%d] %d..%d [%d/%d]\n",
start1, end1, (end1 - start1 + 1), lena,
start2, end2, (end2 - start2 + 1), lenb);
score = embAlignPathCalcSWFast(&p[start1],&q[start2],
end1-start1+1,end2-start2+1,
0,width,
gapopen,gapextend,
path,sub,cvt,
compass,show);
embAlignWalkSWMatrixFast(path,compass,gapopen,gapextend,a,b,
&m,&n,end1-start1+1,end2-start2+1,
0,width,
&start1,&start2);
if(!ajAlignFormatShowsSequences(align))
{
ajAlignDefineCC(align, ajStrGetPtr(m),
ajStrGetPtr(n), ajSeqGetNameC(a),
ajSeqGetNameC(b));
ajAlignSetScoreR(align, score);
}
else
{
embAlignReportLocal(align, a, b,
m,n,start1,start2,
gapopen, gapextend,
score,matrix, begina, beginb);
}
ajAlignWrite(align);
ajAlignReset(align);
ajStrDel(&n);
}
embWordFreeTable(&seq1MatchTable); /* free table of words */
seq1MatchTable=0;
ajStrDel(&m);
}
if(!ajAlignFormatShowsSequences(align))
{
ajMatrixfDel(&matrix);
}
AJFREE(path);
AJFREE(compass);
ajAlignClose(align);
ajAlignDel(&align);
ajSeqallDel(&seq1);
ajSeqDel(&a);
ajSeqsetDel(&seq2);
ajFileClose(&errorf);
embExit();
return 0;
}
/* @prog ssematch ***********************************************************
**
** Searches a DCF file (domain classification file) for secondary structure
** matches.
**
****************************************************************************/
int main(int argc, char **argv)
{
/* Variables declarations */
AjPFile dcfin = NULL; /* Domain classification file */
AjPFile ssin = NULL; /* Secondary structure input file*/
AjPMatrixf matrix = NULL; /* Substitution matrix */
AjPFile out_ss = NULL; /* For ss top matches*/
AjPFile out_se = NULL; /* For se top matches*/
AjPFile outfile = NULL; /* Output file*/
AjPFile logf = NULL; /* Log file */
float gapopen_sss = 0.0; /* Gap insertion penalty */
float gapopen_sse = 0.0;
float gapopen = 0.0;
float gapextend_sss = 0.0; /* Gap extension penalty */
float gapextend_sse = 0.0;
float gapextend = 0.0;
ajint max_hits = 0; /* number of top alignments to display*/
ajint mode = 0;
ajint x = 0;
AjPScop temp_scop = NULL; /* scop object pointer*/
AjPList scop_list = NULL; /* list of scop objects for entire domain
classification file */
AjIList iter = NULL;
AjPStr msg = NULL; /* Pointer to String used for messages */
AjPStr line = NULL;
AjPStr qse = NULL; /* query secondary structure elements*/
AjPStr qss = NULL; /* query secondary structure (by residue)*/
AjPSeq q3se = NULL; /* query secondary structure elements, 3-letter
code*/
AjPSeq q3ss = NULL; /* query secondary structure
(by residue), 3-letter code*/
AjPSeq query = NULL;
/* Read data from acd */
embInitPV("ssematch",argc,argv,"DOMAINATRIX",VERSION);
dcfin = ajAcdGetInfile("dcfinfile");
ssin = ajAcdGetInfile("ssinfile");
max_hits = ajAcdGetInt("maxhits");
matrix = ajAcdGetMatrixf("datafile");
gapopen_sss = ajAcdGetFloat("rgapopen");
gapextend_sss = ajAcdGetFloat("rgapextend");
gapopen_sse = ajAcdGetFloat("egapopen");
gapextend_sse = ajAcdGetFloat("egapextend");
out_ss = ajAcdGetOutfile("outssfile");
out_se = ajAcdGetOutfile("outsefile");
logf = ajAcdGetOutfile("logfile");
/* Create list of scop objects for entire input domain classification file. */
scop_list = ajListNew();
while((temp_scop = (ajScopReadCNew(dcfin, "*"))))
ajListPushAppend(scop_list,temp_scop);
/* Error handing if domain classification file was empty. */
if(!(ajListGetLength(scop_list)))
{
ajWarn("Empty list from scop input file\n");
ajFileClose(&dcfin);
ajFileClose(&ssin);
ajMatrixfDel(&matrix);
ajFileClose(&out_ss);
ajFileClose(&out_se);
ajFileClose(&logf);
while(ajListPop(scop_list, (void *) &temp_scop))
ajScopDel(&temp_scop);
ajListFree(&scop_list);
ajListIterDel(&iter);
ajExit();
return 1;
}
/* Error handling in case of empty query file. */
if(ssin == NULL)
{
ajWarn("Empty secondary structure query file\n");
ajFileClose(&dcfin);
ajFileClose(&ssin);
ajMatrixfDel(&matrix);
ajFileClose(&out_ss);
ajFileClose(&out_se);
ajFileClose(&logf);
while(ajListPop(scop_list, (void *) &temp_scop))
ajScopDel(&temp_scop);
ajListFree(&scop_list);
ajListIterDel(&iter);
ajExit();
return 1;
}
/* Assign sequences in query file to sequence objects. */
qse = ajStrNew();
qss = ajStrNew();
while(ajReadlineTrim(ssin,&line))
{
/* SE string */
if(ajStrPrefixC(line,"SE"))
{
ajFmtScanS(line, "%*s %S", &qse);
/* Convert this string to 3-letter code & then convert to AjPSeq
object. */
q3se = ssematch_convertbases(qse);
}
/* SS string */
else if(ajStrPrefixC(line,"SS"))
{
while((ajReadlineTrim(ssin,&line)) && !ajStrPrefixC(line,"XX"))
ajStrAppendS(&qss,line);
ajStrRemoveWhite(&qss);
/* Convert this string to 3-letter code & then to AjPSeq object. */
q3ss = ssematch_convertbases(qss);
}
}
/* For se & then for ss, modes 0 & 1. */
for(mode = 0; mode <= 1; mode++)
{
/* Assign arguments for alignment function. */
if (mode == 0)
{
query = q3se;
gapopen = gapopen_sse;
gapextend = gapextend_sse;
outfile = out_se;
}
else if(mode == 1)
{
query = q3ss;
gapopen = gapopen_sss;
gapextend = gapextend_sss;
outfile = out_ss;
}
/* Iterate through list of scop objects & calculate alignment scores. */
iter=ajListIterNew(scop_list);
while((temp_scop=(AjPScop)ajListIterGet(iter)))
{
/* The function extracts the se (mode 0) or ss (mode 1) subject
sequences from the scop object, performs a Needleman-Wunsch
global alignment with the query sequence & allocates the score
to the Score element of the scop object*/
if(!(ssematch_NWScore(temp_scop , query, mode, matrix, gapopen, gapextend)))
{
ajFmtPrintF(logf, "%-15s\n", "ALIGNMENT");
ajFmtPrintF(logf, "Could not align sequence in scop domain %S\n ",
temp_scop->Entry);
ajFmtPrintS(&msg, "Could not align sequence in scop domain %S\n ",
temp_scop->Entry);
ajWarn(ajStrGetPtr(msg));
continue;
}
}
ajListIterDel(&iter);
temp_scop = NULL;
/* Sort list of Scop objects by Score */
ajListSort(scop_list, ssematch_CompScoreInv);
iter=ajListIterNew(scop_list);
/* Write top-scoring hits to outfile. */
for(x=0; x < max_hits; x++ )
{
temp_scop=(AjPScop)ajListIterGet(iter);
/* Print score to output file. */
ajFmtPrintF(outfile, "XX ALIGNMENT SCORE %.3f\nXX\n", temp_scop->Score);
/* Could also write alignment - later modification. */
if(!ajScopWrite(outfile, temp_scop))
ajFatal("Could not write output file %S\n", outfile);
}
ajListIterDel(&iter);
temp_scop = NULL;
}
/* Memoryt management. */
ajFileClose(&dcfin);
ajFileClose(&ssin);
ajMatrixfDel(&matrix);
ajFileClose(&out_ss);
ajFileClose(&out_se);
ajFileClose(&logf);
while(ajListPop(scop_list, (void *) &temp_scop))
ajScopDel(&temp_scop);
ajListFree(&scop_list);
ajStrDel(&msg);
ajStrDel(&line);
ajStrDel(&qse);
ajStrDel(&qss);
ajSeqDel(&q3se);
ajSeqDel(&q3ss);
ajExit();
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;
}
/* @prog seqnr **************************************************************
**
** Removes redundancy from DHF files (domain hits files) or other files of
** sequences.
**
****************************************************************************/
int main(int argc, char **argv)
{
/* Variable declarations */
AjPList in = NULL; /* Names of domain hits files (input). */
AjPStr inname = NULL; /* Full name of the current DHF file. */
AjPFile inf = NULL; /* Current DHF file. */
EmbPHitlist infhits = NULL; /* Hitlist from DHF file */
AjBool dosing = ajFalse; /* Filter using singlet sequences. */
AjPDir singlets = NULL; /* Singlets (input). */
AjBool dosets = ajFalse; /* Filter using sets of sequences. */
AjPDir insets = NULL; /* Sets (input). */
AjPStr mode = NULL; /* Mode of operation */
ajint moden = 0; /* Mode 1: single threshold for redundancy
removal, 2: lower and upper thresholds
for redundancy removal. */
float thresh = 0.0; /* Threshold for non-redundancy. */
float threshlow = 0.0; /* Threshold (lower limit). */
float threshup = 0.0; /* Threshold (upper limit). */
AjPMatrixf matrix = NULL; /* Substitution matrix. */
float gapopen = 0.0; /* Gap insertion penalty. */
float gapextend = 0.0; /* Gap extension penalty. */
AjPDirout out = NULL; /* Domain hits files (output). */
AjPFile outf = NULL; /* Current DHF file (output). */
AjBool dored = ajFalse; /* True if redundant hits are output. */
AjPDirout outred = NULL; /* DHF files for redundant hits (output).*/
AjPFile redf = NULL; /* Current DHF file redundancy (output). */
AjPStr outname = NULL; /* Name of output file (re-used). */
AjPFile logf = NULL; /* Log file pointer. */
AjBool ok = ajFalse; /* Housekeeping. */
AjPSeqset seqset = NULL; /* Seqset (re-used). */
AjPSeqin seqin = NULL; /* Seqin (re-used). */
AjPList seq_list = NULL; /* Main list for redundancy removal. */
EmbPDmxNrseq seq_tmp = NULL; /* Temp. pointer for making seq_list. */
ajint seq_siz = 0; /* Size of seq_list. */
AjPUint keep = NULL; /* 1: Sequence in seq_list was classed as
non-redundant, 0: redundant. */
AjPUint nokeep = NULL; /* Inversion of keep array. */
ajint nseqnr = 0; /* No. non-redundant seqs. in seq_list. */
AjPStr filtername= NULL; /* Name of filter file (re-used). */
AjPFile filterf = NULL; /* Current filter file. */
EmbPHitlist hitlist = NULL; /* Hitlist from input file (re-used). */
AjPScopalg scopalg = NULL; /* Scopalg from input file. */
ajint x = 0; /* Housekeeping. */
/* Read data from acd. */
embInitPV("seqnr",argc,argv,"DOMSEARCH",VERSION);
in = ajAcdGetDirlist("dhfinpath");
dosing = ajAcdGetToggle("dosing");
singlets = ajAcdGetDirectory("singletsdir");
dosets = ajAcdGetToggle("dosets");
insets = ajAcdGetDirectory("insetsdir");
mode = ajAcdGetListSingle("mode");
thresh = ajAcdGetFloat("thresh");
threshlow = ajAcdGetFloat("threshlow");
threshup = ajAcdGetFloat("threshup");
matrix = ajAcdGetMatrixf("matrix");
gapopen = ajAcdGetFloat("gapopen");
gapextend = ajAcdGetFloat("gapextend");
out = ajAcdGetOutdir("dhfoutdir");
dored = ajAcdGetToggle("dored");
outred = ajAcdGetOutdir("redoutdir");
logf = ajAcdGetOutfile("logfile");
/* Housekeeping. */
filtername = ajStrNew();
outname = ajStrNew();
if(!(ajStrToInt(mode, &moden)))
ajFatal("Could not parse ACD node option");
/* Process each DHF (input) in turn. */
while(ajListPop(in,(void **)&inname))
{
ajFmtPrint("Processing %S\n", inname);
ajFmtPrintF(logf, "//\n%S\n", inname);
seq_list = ajListNew();
keep = ajUintNew();
nokeep = ajUintNew();
/**********************************/
/* Open DHF file */
/**********************************/
if((inf = ajFileNewInNameS(inname)) == NULL)
ajFatal("Could not open DHF file %S", inname);
/* Read DHF file. */
ok = ajFalse;
if(!(infhits = embHitlistReadFasta(inf)))
{
ajWarn("embHitlistReadFasta call failed in seqnr");
ajFmtPrintF(logf, "embHitlistReadFasta call failed in seqnr\n");
/* Read sequence set instead. */
seqset = ajSeqsetNew();
seqin = ajSeqinNew();
ajSeqinUsa(&seqin, inname);
if(!(ajSeqsetRead(seqset, seqin)))
ajFatal("SeqsetRead failed in seqsearch_psialigned");
if(ajSeqsetGetSize(seqset))
ok = ajTrue;
}
else
if(infhits->N)
ok = ajTrue;
/* Close DHF file. */
ajFileClose(&inf);
/* Process empty DHF files (should never occur). */
if(!ok)
{
ajWarn("Empty input file %S\n", inname);
ajFmtPrintF(logf, "Empty input file %S\n", inname);
if(infhits)
embHitlistDel(&infhits);
if(seqset)
ajSeqsetDel(&seqset);
if(seqin)
ajSeqinDel(&seqin);
continue;
}
/* 1. Create list of sequences from the main input directory.. */
if(infhits)
{
for(x=0; x<infhits->N; x++)
{
AJNEW0(seq_tmp);
seq_tmp->Seq = ajSeqNew();
ajStrAssignS(&seq_tmp->Seq->Acc,infhits->hits[x]->Acc);
ajStrAssignS(&seq_tmp->Seq->Seq,infhits->hits[x]->Seq);
ajListPushAppend(seq_list,seq_tmp);
}
}
else
{
for(x=0;x<ajSeqsetGetSize(seqset);x++)
{
AJNEW0(seq_tmp);
seq_tmp->Seq = ajSeqNew();
ajStrAssignS(&seq_tmp->Seq->Acc, ajSeqsetGetseqAccS(seqset, x));
ajStrAssignS(&seq_tmp->Seq->Seq, ajSeqsetGetseqSeqS(seqset, x));
ajListPushAppend(seq_list,seq_tmp);
}
ajSeqsetDel(&seqset);
ajSeqinDel(&seqin);
}
/**********************************/
/* Open singlets filter file */
/**********************************/
if(dosing)
{
/* Open singlets file. */
ajStrAssignS(&filtername, inname);
ajFilenameTrimPathExt(&filtername);
ajStrInsertS(&filtername, 0, ajDirGetPath(singlets));
ajStrAppendC(&filtername, ".");
ajStrAppendS(&filtername, ajDirGetExt(singlets));
if((filterf = ajFileNewInNameS(filtername)) == NULL)
{
ajWarn("Could not open DHF file %S",
filtername);
ajFmtPrint("Could not open singlets filter file %S",
filtername);
}
else
{
/* Read DHF file. */
ok = ajFalse;
if(!(hitlist = embHitlistReadFasta(filterf)))
{
ajWarn("embHitlistReadFasta call failed in seqnr");
ajFmtPrintF(logf,
"embHitlistReadFasta call failed in seqnr\n");
/* Read sequence set instead. */
seqset = ajSeqsetNew();
seqin = ajSeqinNew();
ajSeqinUsa(&seqin, inname);
if(!(ajSeqsetRead(seqset, seqin)))
ajFatal("SeqsetRead failed in seqnr");
if(ajSeqsetGetSize(seqset))
ok = ajTrue;
}
else
if(hitlist->N)
ok = ajTrue;
/* Close DHF file. */
ajFileClose(&filterf);
/* Process empty DHF files (should never occur). */
if(!ok)
{
ajWarn("Empty singlets filter file %S\n", filtername);
ajFmtPrintF(logf, "Empty singlets filter file %S\n",
filtername);
/* No continue this time. */
}
/* 2. Add sequences from filter directories to List but mark
them up (they are considered in the redundancy calculation
but never appear in the output files). */
if(hitlist)
{
for(x=0; x<hitlist->N; x++)
{
AJNEW0(seq_tmp);
seq_tmp->Seq = ajSeqNew();
seq_tmp->Garbage = ajTrue;
ajStrAssignS(&seq_tmp->Seq->Acc,hitlist->hits[x]->Acc);
ajStrAssignS(&seq_tmp->Seq->Seq,hitlist->hits[x]->Seq);
ajListPushAppend(seq_list,seq_tmp);
}
embHitlistDel(&hitlist);
}
else
{
for(x=0;x<ajSeqsetGetSize(seqset);x++)
{
AJNEW0(seq_tmp);
seq_tmp->Seq = ajSeqNew();
seq_tmp->Garbage = ajTrue;
ajStrAssignS(&seq_tmp->Seq->Acc,
ajSeqsetGetseqAccS(seqset, x));
ajStrAssignS(&seq_tmp->Seq->Seq,
ajSeqsetGetseqSeqS(seqset, x));
ajListPushAppend(seq_list,seq_tmp);
}
ajSeqsetDel(&seqset);
ajSeqinDel(&seqin);
}
}
}
/**********************************/
/* Open sets filter file */
/**********************************/
if(dosets)
{
/* Open sets file. */
ajStrAssignS(&filtername, inname);
ajFilenameTrimPathExt(&filtername);
ajStrInsertS(&filtername, 0, ajDirGetPath(insets));
ajStrAppendC(&filtername, ".");
ajStrAppendS(&filtername, ajDirGetExt(insets));
if((filterf = ajFileNewInNameS(filtername)) == NULL)
{
ajWarn("Could not open DAF file %S", filtername);
ajFmtPrint("Could not open sets filter file %S", filtername);
}
else
{
/* Read DAF file. */
ok = ajFalse;
if(!(ajDmxScopalgRead(filterf, &scopalg)))
{
ajWarn("ajDmxScopalgRead call failed in seqnr");
ajFmtPrintF(logf,
"ajDmxScopalgRead call failed in seqnr\n");
/* Read sequence set instead. */
seqset = ajSeqsetNew();
seqin = ajSeqinNew();
ajSeqinUsa(&seqin, inname);
if(!(ajSeqsetRead(seqset, seqin)))
ajFatal("SeqsetRead failed in seqnr");
if(ajSeqsetGetSize(seqset))
ok = ajTrue;
}
else
if(scopalg->N)
ok = ajTrue;
/* Close DHF file. */
ajFileClose(&filterf);
/* Process empty DHF files (should never occur). */
if(!ok)
{
ajWarn("Empty sets filter file %S\n",
filtername);
ajFmtPrintF(logf, "Empty sets filter file %S\n",
filtername);
/* No continue this time. */
}
/* 2. Add sequences from filter directories to List but mark
them up (they are considered in the redundancy calculation
but never appear in the output files).. */
if(scopalg)
{
for(x=0; x<scopalg->N; x++)
{
AJNEW0(seq_tmp);
seq_tmp->Seq = ajSeqNew();
seq_tmp->Garbage = ajTrue;
ajStrAssignS(&seq_tmp->Seq->Acc,scopalg->Codes[x]);
ajStrAssignS(&seq_tmp->Seq->Seq,scopalg->Seqs[x]);
/* Remove gap char's & whitespace. */
ajStrRemoveGap(&seq_tmp->Seq->Seq);
ajListPushAppend(seq_list,seq_tmp);
}
ajDmxScopalgDel(&scopalg);
}
else
{
for(x=0;x<ajSeqsetGetSize(seqset);x++)
{
AJNEW0(seq_tmp);
seq_tmp->Seq = ajSeqNew();
seq_tmp->Garbage = ajTrue;
ajStrAssignS(&seq_tmp->Seq->Acc,
ajSeqsetGetseqAccS(seqset, x));
ajStrAssignS(&seq_tmp->Seq->Seq,
ajSeqsetGetseqSeqS(seqset, x));
ajListPushAppend(seq_list,seq_tmp);
}
ajSeqsetDel(&seqset);
ajSeqinDel(&seqin);
}
}
}
/* 4. Identify redundant domains.. */
if(moden == 1)
{
if((!embDmxSeqNR(seq_list, &keep, &nseqnr, matrix, gapopen,
gapextend, thresh, ajTrue)))
ajFatal("embDmxSeqNR failure in seqnr");
}
else
{
if((!embDmxSeqNRRange(seq_list, &keep, &nseqnr, matrix, gapopen,
gapextend, threshlow, threshup, ajTrue)))
ajFatal("embDmxSeqNR failure in seqnr");
}
seq_siz = ajListGetLength(seq_list);
for(x=0; x<seq_siz; x++)
if(ajUintGet(keep, x) == 1)
ajUintPut(&nokeep, x, 0);
else
ajUintPut(&nokeep, x, 1);
/* Create output files. */
ajStrAssignS(&outname, inname);
ajFilenameTrimPathExt(&outname);
outf = ajFileNewOutNameDirS(outname, out);
if(dored)
redf = ajFileNewOutNameDirS(outname, outred);
/* 5. Write non-redundant domains to main output directory.
6. If specified, write redundant domains to output directory. */
embHitlistWriteSubsetFasta(outf, infhits, keep);
if(dored)
embHitlistWriteSubsetFasta(redf, infhits, nokeep);
embHitlistDel(&infhits);
ajFileClose(&outf);
ajFileClose(&redf);
ajStrDel(&inname);
while(ajListPop(seq_list, (void **) &seq_tmp))
{
ajSeqDel(&seq_tmp->Seq);
AJFREE(seq_tmp);
}
ajListFree(&seq_list);
ajUintDel(&keep);
ajUintDel(&nokeep);
}
/* Tidy up. */
ajListFree(&in);
if(singlets)
ajDirDel(&singlets);
if(insets)
ajDirDel(&insets);
ajDiroutDel(&out);
if(outred)
ajDiroutDel(&outred);
ajFileClose(&logf);
ajMatrixfDel(&matrix);
ajStrDel(&filtername);
ajStrDel(&outname);
ajStrDel(&mode);
embExit();
return 0;
}