static void
onefetch_subseq(ESL_GETOPTS *go, FILE *ofp, ESL_SQFILE *sqfp, char *newname, char *key, uint32_t given_start, uint32_t given_end)
{
int start, end;
int do_revcomp;
ESL_SQ *sq = esl_sq_Create();
if (sqfp->data.ascii.ssi == NULL) esl_fatal("no ssi index");
/* reverse complement indicated by coords. */
/* -c 52: would be 52,0, so watch out for given_end = 0 case */
if (given_end != 0 && given_start > given_end)
{ start = given_end; end = given_start; do_revcomp = TRUE; }
else
{ start = given_start; end = given_end; do_revcomp = FALSE; }
if (esl_sqio_FetchSubseq(sqfp, key, start, end, sq) != eslOK) esl_fatal(esl_sqfile_GetErrorBuf(sqfp));
if (newname != NULL) esl_sq_SetName(sq, newname);
else esl_sq_FormatName(sq, "%s/%d-%d", key, given_start, (given_end == 0) ? sq->L : given_end);
/* Two ways we might have been asked to revcomp: by coord, or by -r option */
/* (If both happen, they'll cancel each other out) */
if (do_revcomp)
if (esl_sq_ReverseComplement(sq) != eslOK) esl_fatal("Failed to reverse complement %s; is it a protein?\n", sq->name);
if (esl_opt_GetBoolean(go, "-r"))
if (esl_sq_ReverseComplement(sq) != eslOK) esl_fatal("Failed to reverse complement %s; is it a protein?\n", sq->name);
esl_sqio_Write(ofp, sq, eslSQFILE_FASTA, FALSE);
esl_sq_Destroy(sq);
}
/* onefetch():
* Given one <key> (a seq name or accession), retrieve the corresponding sequence.
* In SSI mode, we can do this quickly by positioning the file, then regurgitating
* every line until the end-of-record marker; we don't even have to parse.
* Without an SSI index, we have to parse the file sequentially 'til we find
* the one we're after.
*/
static void
onefetch(ESL_GETOPTS *go, FILE *ofp, char *key, ESL_SQFILE *sqfp)
{
ESL_SQ *sq = esl_sq_Create();
int do_revcomp = esl_opt_GetBoolean(go, "-r");
char *newname = esl_opt_GetString(go, "-n");
int status;
/* Try to position the file at the desired sequence with SSI. */
if (sqfp->data.ascii.ssi != NULL)
{
status = esl_sqfile_PositionByKey(sqfp, key);
if (status == eslENOTFOUND) esl_fatal("seq %s not found in SSI index for file %s\n", key, sqfp->filename);
else if (status == eslEFORMAT) esl_fatal("Failed to parse SSI index for %s\n", sqfp->filename);
else if (status != eslOK) esl_fatal("Failed to look up location of seq %s in SSI index of file %s\n", key, sqfp->filename);
status = esl_sqio_Read(sqfp, sq);
if (status == eslEFORMAT) esl_fatal("Parse failed (sequence file %s):\n%s\n",
sqfp->filename, esl_sqfile_GetErrorBuf(sqfp));
else if (status == eslEOF) esl_fatal("Unexpected EOF reading sequence file %s",
status, sqfp->filename);
else if (status != eslOK) esl_fatal("Unexpected error %d reading sequence file %s",
status, sqfp->filename);
if (strcmp(key, sq->name) != 0 && strcmp(key, sq->acc) != 0)
esl_fatal("whoa, internal error; found the wrong sequence %s, not %s", sq->name, key);
}
else
{ /* Else, we have to read the whole damn file sequentially until we find the seq */
while ((status = esl_sqio_Read(sqfp, sq)) != eslEOF) {
if (status == eslEFORMAT) esl_fatal("Parse failed (sequence file %s):\n%s\n",
sqfp->filename, esl_sqfile_GetErrorBuf(sqfp));
else if (status != eslOK) esl_fatal("Unexpected error %d reading sequence file %s",
status, sqfp->filename);
if (strcmp(key, sq->name) == 0 || strcmp(key, sq->acc) == 0) break;
esl_sq_Reuse(sq);
}
if (status == eslEOF) esl_fatal("Failed to find sequence %s in file %s\n", key, sqfp->filename);
}
if (do_revcomp == FALSE && newname == NULL && ! esl_sqio_IsAlignment(sqfp->format))
{ /* If we're not manipulating the sequence in any way, and it's not from an alignment file, we can Echo() it. */
if (esl_sqio_Echo(sqfp, sq, ofp) != eslOK) esl_fatal("Echo failed: %s\n", esl_sqfile_GetErrorBuf(sqfp));
}
else
{ /* Otherwise we Write() the parsed version. */
if (do_revcomp && esl_sq_ReverseComplement(sq) != eslOK) esl_fatal("Failed to reverse complement %s; is it a protein?\n", sq->name);
if (newname != NULL) esl_sq_SetName(sq, newname);
esl_sqio_Write(ofp, sq, eslSQFILE_FASTA, FALSE);
}
esl_sq_Destroy(sq);
}
/* seq_generation()
*
* Generating sequences.
*/
static int
seq_generation(ESL_GETOPTS *go, ESL_RANDOMNESS *r, FILE *ofp, int outfmt)
{
ESL_ALPHABET *abc = NULL;
ESL_SQ *sq = NULL;
double *fq = NULL;
int alphatype = eslUNKNOWN; // static checkers can't see that 1 of --rna, --dna, --amino must be true
int N = esl_opt_GetInteger(go, "-N");
int L = esl_opt_GetInteger(go, "-L");
int i;
int status;
if (L <= 0) esl_fatal("To generate sequences, set -L option (length of generated seqs) > 0 ");
if (esl_opt_GetBoolean(go, "--rna")) alphatype = eslRNA;
if (esl_opt_GetBoolean(go, "--dna")) alphatype = eslDNA;
if (esl_opt_GetBoolean(go, "--amino")) alphatype = eslAMINO;
abc = esl_alphabet_Create(alphatype);
sq = esl_sq_CreateDigital(abc);
esl_sq_GrowTo(sq, L);
/* Pick the iid frequency distribution to use */
ESL_ALLOC(fq, sizeof(double) * abc->K);
switch (alphatype) {
case eslRNA:
case eslDNA: esl_vec_DSet(fq, 4, 0.25); break;
case eslAMINO: esl_composition_SW34(fq); break;
default: esl_vec_DSet(fq, abc->K, 1.0 / (double) abc->K); break;
}
/* generate */
for (i = 0; i < N; i++)
{
esl_rsq_xIID(r, fq, abc->K, L, sq->dsq);
if (N > 1) esl_sq_FormatName(sq, "random%d", i);
else esl_sq_SetName(sq, "random");
sq->n = L;
esl_sqio_Write(ofp, sq, outfmt, FALSE);
}
free(fq);
esl_alphabet_Destroy(abc);
esl_sq_Destroy(sq);
return eslOK;
ERROR:
if (fq != NULL) free(fq);
esl_alphabet_Destroy(abc);
esl_sq_Destroy(sq);
return status;
}
int
main(int argc, char **argv)
{
ESL_GETOPTS *go = esl_getopts_CreateDefaultApp(options, 1, argc, argv, banner, usage);
char *hmmfile = esl_opt_GetArg(go, 1);
int N = esl_opt_GetInteger(go, "-N");
ESL_STOPWATCH *w = esl_stopwatch_Create();
ESL_RANDOMNESS *r = esl_randomness_CreateFast(esl_opt_GetInteger(go, "-s"));
ESL_ALPHABET *abc = NULL;
P7_HMMFILE *hfp = NULL;
P7_HMM *hmm = NULL;
P7_BG *bg = NULL;
P7_PROFILE *gm = NULL;
P7_OPROFILE *om = NULL;
P7_TRACE *tr = NULL;
ESL_SQ *sq = NULL;
P7_ALIDISPLAY *ad = NULL;
int i,z;
if (p7_hmmfile_Open(hmmfile, NULL, &hfp) != eslOK) p7_Fail("Failed to open HMM file %s", hmmfile);
if (p7_hmmfile_Read(hfp, &abc, &hmm) != eslOK) p7_Fail("Failed to read HMM");
p7_hmmfile_Close(hfp);
bg = p7_bg_Create(abc);
p7_bg_SetLength(bg, 0);
gm = p7_profile_Create(hmm->M, abc);
p7_ProfileConfig(hmm, bg, gm, 0, p7_UNIGLOCAL); /* that sets N,C,J to generate nothing */
om = p7_oprofile_Create(gm->M, abc);
p7_oprofile_Convert(gm, om);
if (esl_opt_GetBoolean(go, "-p")) tr = p7_trace_CreateWithPP();
else tr = p7_trace_Create();
sq = esl_sq_CreateDigital(abc);
esl_stopwatch_Start(w);
for (i = 0; i < N; i++)
{
p7_ProfileEmit(r, hmm, gm, bg, sq, tr);
esl_sq_SetName(sq, "random");
if (! esl_opt_GetBoolean(go, "-b"))
{
if (esl_opt_GetBoolean(go, "-p"))
for (z = 0; z < tr->N; z++)
if (tr->i[z] > 0) tr->pp[z] = esl_random(r);
ad = p7_alidisplay_Create(tr, 0, om, sq);
p7_alidisplay_Print(stdout, ad, 40, 80, FALSE);
p7_alidisplay_Destroy(ad);
}
p7_trace_Reuse(tr);
esl_sq_Reuse(sq);
}
esl_stopwatch_Stop(w);
esl_stopwatch_Display(stdout, w, "# CPU time: ");
esl_sq_Destroy(sq);
p7_trace_Destroy(tr);
p7_oprofile_Destroy(om);
p7_profile_Destroy(gm);
p7_bg_Destroy(bg);
p7_hmm_Destroy(hmm);
esl_alphabet_Destroy(abc);
esl_randomness_Destroy(r);
esl_stopwatch_Destroy(w);
esl_getopts_Destroy(go);
return 0;
}
