static int
process_commandline(int argc, char **argv, ESL_GETOPTS **ret_go, char **ret_fmfile, char **ret_qfile)
{
ESL_GETOPTS *go = esl_getopts_Create(options);
int status;
if (esl_opt_ProcessEnvironment(go) != eslOK) { if (printf("Failed to process environment: %s\n", go->errbuf) < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "write failed"); goto FAILURE; }
if (esl_opt_ProcessCmdline(go, argc, argv) != eslOK) { if (printf("Failed to parse command line: %s\n", go->errbuf) < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "write failed"); goto FAILURE; }
if (esl_opt_VerifyConfig(go) != eslOK) { if (printf("Failed to parse command line: %s\n", go->errbuf) < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "write failed"); goto FAILURE; }
/* help format: */
if (esl_opt_GetBoolean(go, "-h") == TRUE)
{
esl_banner(stdout, argv[0], banner);
esl_usage(stdout, argv[0], usage);
if (puts("\nBasic options:") < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "write failed");
esl_opt_DisplayHelp(stdout, go, 1, 2, 80); /* 1= group; 2 = indentation; 120=textwidth*/
if (puts("\nSpecial options:") < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "write failed");
esl_opt_DisplayHelp(stdout, go, 2, 2, 80); /* 2= group; 2 = indentation; 120=textwidth*/
exit(0);
}
if (esl_opt_ArgNumber(go) != 2) { if (puts("Incorrect number of command line arguments.") < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "write failed"); goto FAILURE; }
if ((*ret_qfile = esl_opt_GetArg(go, 1)) == NULL) { if (puts("Failed to get <qfile> argument on command line") < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "write failed"); goto FAILURE; }
if ((*ret_fmfile = esl_opt_GetArg(go, 2)) == NULL) { if (puts("Failed to get <fmfile> argument on command line") < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "write failed"); goto FAILURE; }
/* Validate any attempted use of stdin streams */
if (esl_strcmp(*ret_fmfile, "-") == 0 && esl_strcmp(*ret_qfile, "-") == 0)
{ if (puts("Either <fmfile> or <qfile> may be '-' (to read from stdin), but not both.") < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "write failed"); goto FAILURE; }
*ret_go = go;
return eslOK;
FAILURE: /* all errors handled here are user errors, so be polite. */
esl_usage(stdout, argv[0], usage);
puts("\nwhere basic options are:");
esl_opt_DisplayHelp(stdout, go, 1, 2, 80); /* 1= group; 2 = indentation; 80=textwidth*/
printf("\nTo see more help on available options, do %s -h\n\n", argv[0]);
esl_getopts_Destroy(go);
exit(1);
ERROR:
if (go) esl_getopts_Destroy(go);
exit(status);
}
/* process_commandline()
*
* Processes the commandline, filling in fields in <cfg> and creating and returning
* an <ESL_GETOPTS> options structure. The help page (hmmsearch -h) is formatted
* here.
*/
static int
process_commandline(int argc, char **argv, ESL_GETOPTS **ret_go, char **ret_hmmfile, char **ret_seqfile)
{
ESL_GETOPTS *go = esl_getopts_Create(options);
int status;
if (esl_opt_ProcessEnvironment(go) != eslOK) { if (printf("Failed to process environment: %s\n", go->errbuf) < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "write failed"); goto FAILURE; }
if (esl_opt_ProcessCmdline(go, argc, argv) != eslOK) { if (printf("Failed to parse command line: %s\n", go->errbuf) < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "write failed"); goto FAILURE; }
if (esl_opt_VerifyConfig(go) != eslOK) { if (printf("Failed to parse command line: %s\n", go->errbuf) < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "write failed"); goto FAILURE; }
/* help format: */
if (esl_opt_GetBoolean(go, "-h") == TRUE)
{
p7_banner(stdout, argv[0], banner);
esl_usage(stdout, argv[0], usage);
if (puts("\nBasic options:") < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "write failed");
esl_opt_DisplayHelp(stdout, go, 1, 2, 80); /* 1= group; 2 = indentation; 80=textwidth*/
if (puts("\nOptions controlling output:") < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "write failed");
esl_opt_DisplayHelp(stdout, go, 2, 2, 80);
if (puts("\nOptions controlling reporting thresholds:") < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "write failed");
esl_opt_DisplayHelp(stdout, go, 4, 2, 80);
if (puts("\nOptions controlling inclusion (significance) thresholds:") < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "write failed");
esl_opt_DisplayHelp(stdout, go, 5, 2, 80);
if (puts("\nOptions for model-specific thresholding:") < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "write failed");
esl_opt_DisplayHelp(stdout, go, 6, 2, 80);
if (puts("\nOptions controlling acceleration heuristics:") < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "write failed");
esl_opt_DisplayHelp(stdout, go, 7, 2, 80);
if (puts("\nOther expert options:") < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "write failed");
esl_opt_DisplayHelp(stdout, go, 12, 2, 80);
exit(0);
}
if (esl_opt_ArgNumber(go) != 2) { if (puts("Incorrect number of command line arguments.") < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "write failed"); goto FAILURE; }
if ((*ret_hmmfile = esl_opt_GetArg(go, 1)) == NULL) { if (puts("Failed to get <hmmdb> argument on command line") < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "write failed"); goto FAILURE; }
if ((*ret_seqfile = esl_opt_GetArg(go, 2)) == NULL) { if (puts("Failed to get <seqfile> argument on command line") < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "write failed"); goto FAILURE; }
/* Validate any attempted use of stdin streams */
if (strcmp(*ret_hmmfile, "-") == 0)
{ if (puts("nhmmscan cannot read <hmm database> from stdin stream, because it must have hmmpress'ed auxfiles") < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "write failed"); goto FAILURE; }
*ret_go = go;
return eslOK;
FAILURE: /* all errors handled here are user errors, so be polite. */
esl_usage(stdout, argv[0], usage);
if (puts("\nwhere most common options are:") < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "write failed");
esl_opt_DisplayHelp(stdout, go, 1, 2, 80); /* 1= group; 2 = indentation; 80=textwidth*/
if (printf("\nTo see more help on available options, do %s -h\n\n", argv[0]) < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "write failed");
esl_getopts_Destroy(go);
exit(1);
ERROR:
if (go) esl_getopts_Destroy(go);
exit(status);
}
static int
output_header(FM_METADATA *meta, FILE *ofp, const ESL_GETOPTS *go, char *fmfile, char *qfile)
{
char *alph;
char *appname = NULL;
int status;
if (meta->alph_type == fm_DNA) alph = "dna";
else if (meta->alph_type == fm_AMINO) alph = "amino";
if ((status = esl_FileTail(go->argv[0], FALSE, &appname)) != eslOK) return status;
if (fprintf(ofp, "# %s :: %s\n", appname, banner) < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "write failed");
if (fprintf(ofp, "# %s\n", EASEL_COPYRIGHT) < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "write failed");
if (fprintf(ofp, "# %s\n", EASEL_LICENSE) < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "write failed");
if (fprintf(ofp, "# - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -\n") < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "write failed");
if (fprintf(ofp, "# input binary-formatted HMMER database: %s\n", fmfile) < 0) ESL_EXCEPTION_SYS(eslEWRITE, "write failed");
if (fprintf(ofp, "# input file of query sequences: %s\n", qfile) < 0) ESL_EXCEPTION_SYS(eslEWRITE, "write failed");
if (esl_opt_IsUsed(go, "--out")) {
char *outfile = esl_opt_GetString(go, "--out");
if (fprintf(ofp, "# output file containing list of hits: %s\n", (esl_strcmp(outfile, "-") == 0 ? "stdout" : outfile)) < 0) ESL_EXCEPTION_SYS(eslEWRITE, "write failed");
}
if (esl_opt_IsUsed(go, "--count_only") && fprintf(ofp, "# output only counts, not hit locations\n") < 0) ESL_EXCEPTION_SYS(eslEWRITE, "write failed");
if (fprintf(ofp, "# alphabet : %s\n", alph) < 0) ESL_EXCEPTION_SYS(eslEWRITE, "write failed");
if (fprintf(ofp, "# bin_length : %d\n", meta->freq_cnt_b) < 0) ESL_EXCEPTION_SYS(eslEWRITE, "write failed");
if (fprintf(ofp, "# suffix array sample rate: %d\n", meta->freq_SA) < 0) ESL_EXCEPTION_SYS(eslEWRITE, "write failed");
if (fprintf(ofp, "# - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -\n\n") < 0) ESL_EXCEPTION_SYS(eslEWRITE, "write failed");
if (appname) free(appname);
return eslOK;
ERROR:
if (appname) free(appname);
return status;
}
static int
process_commandline(int argc, char **argv, ESL_GETOPTS **ret_go, char **ret_alifile, char **ret_postalifile)
{
ESL_GETOPTS *go = esl_getopts_Create(options);
int status;
if (esl_opt_ProcessEnvironment(go) != eslOK) { if (printf("Failed to process environment:\n%s\n", go->errbuf) < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "write failed"); goto FAILURE; }
if (esl_opt_ProcessCmdline(go, argc, argv) != eslOK) { if (printf("Failed to parse command line:\n%s\n", go->errbuf) < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "write failed"); goto FAILURE; }
if (esl_opt_VerifyConfig(go) != eslOK) { if (printf("Failed to parse command line:\n%s\n", go->errbuf) < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "write failed"); goto FAILURE; }
/* help format: */
if (esl_opt_GetBoolean(go, "-h") == TRUE)
{
p7_banner(stdout, argv[0], banner);
esl_usage(stdout, argv[0], usage);
if (puts("\nBasic options:") < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "write failed");
esl_opt_DisplayHelp(stdout, go, 1, 2, 80);
if (puts("\nMask range options (format: --xxx 10-20,30-40 ) :") < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "write failed");
esl_opt_DisplayHelp(stdout, go, 5, 2, 80);
if (puts("\nOptions for selecting alphabet rather than guessing it:") < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "write failed");
esl_opt_DisplayHelp(stdout, go, 2, 2, 80);
if (puts("\nAlternative model construction strategies:") < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "write failed");
esl_opt_DisplayHelp(stdout, go, 3, 2, 80);
if (puts("\nAlternative relative sequence weighting strategies:") < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "write failed");
esl_opt_DisplayHelp(stdout, go, 4, 2, 80);
if (puts("\nOther options:") < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "write failed");
esl_opt_DisplayHelp(stdout, go, 8, 2, 80);
exit(0);
}
if (esl_opt_ArgNumber(go) > 2) { if (puts("Incorrect number of command line arguments.") < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "write failed"); goto FAILURE; }
if ((*ret_alifile = esl_opt_GetArg(go, 1)) == NULL) { if (puts("Failed to get <msafile> argument on command line") < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "write failed"); goto FAILURE; }
if (esl_opt_IsUsed(go, "--alirange") || esl_opt_IsUsed(go, "--modelrange") ) {
if ((*ret_postalifile = esl_opt_GetArg(go, 2)) == NULL) { if (puts("Failed to get <postmsafile> argument on command line") < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "write failed"); goto FAILURE; }
}
if (strcmp(*ret_alifile, "-") == 0 && ! esl_opt_IsOn(go, "--informat"))
{ if (puts("Must specify --informat to read <alifile> from stdin ('-')") < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "write failed"); goto FAILURE; }
*ret_go = go;
return eslOK;
FAILURE: /* all errors handled here are user errors, so be polite. */
esl_usage(stdout, argv[0], usage);
puts("\nwhere basic options are:");
esl_opt_DisplayHelp(stdout, go, 1, 2, 80);
printf("\nTo see more help on other available options, do:\n %s -h\n\n", argv[0]);
esl_getopts_Destroy(go);
exit(1);
ERROR:
if (go) esl_getopts_Destroy(go);
exit(status);
}
int
main(int argc, char **argv)
{
int i,j;
ESL_GETOPTS *go = NULL; /* command line processing */
ESL_STOPWATCH *w = esl_stopwatch_Create();
int status;
ESL_MSA *msa = NULL;
FILE *ofp = NULL; /* output file (default is stdout) */
ESL_ALPHABET *abc = NULL; /* digital alphabet */
char *alifile; /* name of the alignment file we're building HMMs from */
ESLX_MSAFILE *afp = NULL; /* open alifile */
int infmt = eslMSAFILE_UNKNOWN; /* autodetect alignment format by default. */
int outfmt = eslMSAFILE_STOCKHOLM;
char *postmsafile; /* optional file to resave annotated, modified MSAs to */
FILE *postmsafp = NULL; /* open <postmsafile>, or NULL */
int mask_range_cnt = 0;
uint32_t mask_starts[100]; // over-the-top allocation.
uint32_t mask_ends[100];
char *rangestr;
char *range;
int *map = NULL; /* map[i]=j, means model position i comes from column j of the alignment; 1..alen */
int keep_mm;
/* Set processor specific flags */
impl_Init();
alifile = NULL;
postmsafile = NULL;
/* Parse the command line
*/
process_commandline(argc, argv, &go, &alifile, &postmsafile);
keep_mm = esl_opt_IsUsed(go, "--apendmask");
/* Initialize what we can in the config structure (without knowing the alphabet yet).
* Fields controlled by masters are set up in usual_master() or mpi_master()
* Fields used by workers are set up in mpi_worker()
*/
ofp = NULL;
infmt = eslMSAFILE_UNKNOWN;
afp = NULL;
abc = NULL;
if (esl_opt_IsOn(go, "--informat")) {
infmt = eslx_msafile_EncodeFormat(esl_opt_GetString(go, "--informat"));
if (infmt == eslMSAFILE_UNKNOWN) p7_Fail("%s is not a recognized input sequence file format\n", esl_opt_GetString(go, "--informat"));
}
/* Determine output alignment file format */
outfmt = eslx_msafile_EncodeFormat(esl_opt_GetString(go, "--outformat"));
if (outfmt == eslMSAFILE_UNKNOWN) p7_Fail(argv[0], "%s is not a recognized output MSA file format\n", esl_opt_GetString(go, "--outformat"));
/* Parse the ranges */
if (esl_opt_IsUsed(go, "--alirange")) {
esl_strdup(esl_opt_GetString(go, "--alirange"), -1, &rangestr) ;
} else if (esl_opt_IsUsed(go, "--modelrange")) {
esl_strdup(esl_opt_GetString(go, "--modelrange"), -1, &rangestr) ;
} else if (esl_opt_IsUsed(go, "--model2ali")) {
esl_strdup(esl_opt_GetString(go, "--model2ali"), -1, &rangestr) ;
} else if (esl_opt_IsUsed(go, "--ali2model")) {
esl_strdup(esl_opt_GetString(go, "--ali2model"), -1, &rangestr) ;
} else {
if (puts("Must specify mask range with --modelrange, --alirange, --model2ali, or --ali2model\n") < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "write failed"); goto ERROR;
}
while ( (status = esl_strtok(&rangestr, ",", &range) ) == eslOK) {
status = esl_regexp_ParseCoordString(range, mask_starts + mask_range_cnt, mask_ends + mask_range_cnt );
if (status == eslESYNTAX) esl_fatal("range flags take coords <from>..<to>; %s not recognized", range);
if (status == eslFAIL) esl_fatal("Failed to find <from> or <to> coord in %s", range);
mask_range_cnt++;
}
/* Start timing. */
esl_stopwatch_Start(w);
/* Open files, set alphabet.
* afp - open alignment file for input
* abc - alphabet expected or guessed in ali file
* postmsafp - open MSA output file
* ofp - optional open output file, or stdout
*/
if (esl_opt_GetBoolean(go, "--amino")) abc = esl_alphabet_Create(eslAMINO);
else if (esl_opt_GetBoolean(go, "--dna")) abc = esl_alphabet_Create(eslDNA);
else if (esl_opt_GetBoolean(go, "--rna")) abc = esl_alphabet_Create(eslRNA);
else abc = NULL;
status = eslx_msafile_Open(&abc, alifile, NULL, infmt, NULL, &afp);
if (status != eslOK) eslx_msafile_OpenFailure(afp, status);
if (esl_opt_IsUsed(go, "--alirange") || esl_opt_IsUsed(go, "--modelrange") ) {
postmsafp = fopen(postmsafile, "w");
if (postmsafp == NULL) p7_Fail("Failed to MSA output file %s for writing", postmsafile);
}
if (esl_opt_IsUsed(go, "-o"))
{
ofp = fopen(esl_opt_GetString(go, "-o"), "w");
if (ofp == NULL) p7_Fail("Failed to open -o output file %s\n", esl_opt_GetString(go, "-o"));
}
else ofp = stdout;
/* Looks like the i/o is set up successfully...
* Initial output to the user
*/
output_header(go, ofp, alifile, postmsafile); /* cheery output header */
/* read the alignment */
if ((status = eslx_msafile_Read(afp, &msa)) != eslOK) eslx_msafile_ReadFailure(afp, status);
if (esl_opt_IsUsed(go, "--alirange") || esl_opt_IsUsed(go, "--modelrange") ) {
/* add/modify mmline for the mask */
if (msa->mm == NULL) {
ESL_ALLOC(msa->mm, msa->alen);
keep_mm = FALSE;
}
if (!keep_mm)
for (i=0; i<msa->alen; i++) msa->mm[i] = '.';
}
// convert model coordinates to alignment coordinates, if necessary
if (esl_opt_IsUsed(go, "--modelrange") || esl_opt_IsUsed(go, "--model2ali") || esl_opt_IsUsed(go, "--ali2model") ) {
float symfrac = esl_opt_GetReal(go, "--symfrac");
int do_hand = esl_opt_IsOn(go, "--hand");
int L;
//same as p7_builder relative_weights
if (esl_opt_IsOn(go, "--wnone") ) { esl_vec_DSet(msa->wgt, msa->nseq, 1.); }
else if (esl_opt_IsOn(go, "--wgiven") ) ;
else if (esl_opt_IsOn(go, "--wpb") ) status = esl_msaweight_PB(msa);
else if (esl_opt_IsOn(go, "--wgsc") ) status = esl_msaweight_GSC(msa);
else if (esl_opt_IsOn(go, "--wblosum")) status = esl_msaweight_BLOSUM(msa, esl_opt_GetReal(go, "--wid"));
if ((status = esl_msa_MarkFragments(msa, esl_opt_GetReal(go, "--fragthresh"))) != eslOK) goto ERROR;
//build a map of model mask coordinates to alignment coords
ESL_ALLOC(map, sizeof(int) * (msa->alen+1));
L = p7_Alimask_MakeModel2AliMap(msa, do_hand, symfrac, map );
if ( esl_opt_IsUsed(go, "--model2ali") ) {
//print mapping
printf ("model coordinates alignment coordinates\n");
for (i=0; i<mask_range_cnt; i++)
printf ("%8d..%-8d -> %8d..%-8d\n", mask_starts[i], mask_ends[i], map[mask_starts[i]-1], map[mask_ends[i]-1]);
/* If I wanted to, I could print all the map values independently:
printf("\n\n-----------\n");
printf("Map\n");
printf("---\n");
for (i=0; i<L; i++)
printf("%d -> %d\n", i+1, map[i]);
*/
} else if ( esl_opt_IsUsed(go, "--ali2model") ) {
//print mapping (requires scanning the inverted map
int alistart = 0;
int aliend = 0;
printf ("alignment coordinates model coordinates\n");
for (i=0; i<mask_range_cnt; i++) {
//find j for ali positions
while (map[alistart] < mask_starts[i] )
alistart++;
aliend = alistart;
while (map[aliend] < mask_ends[i] )
aliend++;
printf (" %8d..%-8d -> %8d..%-8d\n", map[alistart], map[aliend], alistart+1, aliend+1);
}
} else {
//convert the mask coords based on map
for (i=0; i<mask_range_cnt; i++) {
mask_starts[i] = map[mask_starts[i]-1]; //-1 because mmline is offset by one relative to the 1-base alignment
mask_ends[i] = map[mask_ends[i]-1];
}
}
}
if (esl_opt_IsUsed(go, "--alirange") || esl_opt_IsUsed(go, "--modelrange") ) {
//overwrite '.' with 'm' everywhere the range says to do it
for (i=0; i<mask_range_cnt; i++)
for (j=mask_starts[i]; j<=mask_ends[i]; j++)
msa->mm[j-1] = 'm';
if ((status = eslx_msafile_Write(postmsafp, msa, outfmt)) != eslOK) ESL_XEXCEPTION_SYS(eslEWRITE, "write failed");
}
esl_stopwatch_Stop(w);
if (esl_opt_IsOn(go, "-o")) fclose(ofp);
if (postmsafp) fclose(postmsafp);
if (afp) eslx_msafile_Close(afp);
if (abc) esl_alphabet_Destroy(abc);
esl_getopts_Destroy(go);
esl_stopwatch_Destroy(w);
return 0;
ERROR:
return eslFAIL;
}
/* Function: esl_msafile_a2m_Write()
* Synopsis: Write an A2M (UCSC SAM) dotless format alignment to a stream.
*
* Purpose: Write alignment <msa> in dotless UCSC A2M format to a
* stream <fp>.
*
* The <msa> should have a valid reference line <msa->rf>,
* with alphanumeric characters marking consensus (match)
* columns, and non-alphanumeric characters marking
* nonconsensus (insert) columns. If it does not,
* then as a fallback, the first sequence in the alignment is
* considered to be the consensus.
*
* In "dotless" A2M format, gap characters (.) in insert
* columns are omitted; therefore sequences can be of
* different lengths, but each sequence has the same number
* of consensus columns (residue or -).
*
* A2M format cannot represent missing data symbols
* (Easel's ~). Any missing data symbols are converted to
* gaps.
*
* A2M format cannot represent pyrrolysine residues in
* amino acid sequences, because it treats 'O' symbols
* specially, as indicating a position at which a
* free-insertion module (FIM) should be created. Any 'O'
* in the <msa> is written instead as an unknown
* residue ('X', in protein sequences).
*
* Args: fp - open output stream
* msa - MSA to write
*
* Returns: <eslOK> on success.
*
* Throws: <eslEMEM> on allocation failure.
* <eslEWRITE> on any system write error, such as filled disk.
*/
int
esl_msafile_a2m_Write(FILE *fp, const ESL_MSA *msa)
{
char *buf = NULL;
int cpl = 60;
int bpos;
int pos;
int is_consensus;
int is_residue;
int do_dotless = TRUE; /* just changing this to FALSE makes it write dots too */
int i;
int sym;
int status;
ESL_ALLOC(buf, sizeof(char) * (cpl+1));
for (i = 0; i < msa->nseq; i++)
{
/* Construct the name/description line */
if (fprintf(fp, ">%s", msa->sqname[i]) < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "a2m msa file write failed");
if (msa->sqacc != NULL && msa->sqacc[i] != NULL) { if (fprintf(fp, " %s", msa->sqacc[i]) < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "a2m msa file write failed"); }
if (msa->sqdesc != NULL && msa->sqdesc[i] != NULL) { if (fprintf(fp, " %s", msa->sqdesc[i]) < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "a2m msa file write failed"); }
if (fputc('\n', fp) < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "a2m msa file write failed");
#ifdef eslAUGMENT_ALPHABET
if (msa->abc)
{
pos = 0;
while (pos < msa->alen)
{
for (bpos = 0; pos < msa->alen && bpos < cpl; pos++)
{
sym = msa->abc->sym[msa->ax[i][pos+1]]; /* note off-by-one in digitized aseq: 1..alen */
is_residue = esl_abc_XIsResidue(msa->abc, msa->ax[i][pos+1]);
if (msa->rf) is_consensus = (isalnum(msa->rf[pos]) ? TRUE : FALSE);
else is_consensus = (esl_abc_XIsResidue(msa->abc, msa->ax[0][pos+1]) ? TRUE : FALSE);
if (sym == 'O') sym = esl_abc_XGetUnknown(msa->abc); /* watch out: O means "insert a FIM" in a2m format, not pyrrolysine */
if (is_consensus) { buf[bpos++] = (is_residue ? toupper(sym) : '-'); }
else if (is_residue) { buf[bpos++] = tolower(sym); }
else if (! do_dotless) { buf[bpos++] = '.'; }
}
buf[bpos] = '\0';
if (bpos) { if (fprintf(fp, "%s\n", buf) < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "a2m msa file write failed");}
}
}
#endif
if (! msa->abc)
{
pos = 0;
while (pos < msa->alen)
{
for (bpos = 0; pos < msa->alen && bpos < cpl; pos++)
{
sym = msa->aseq[i][pos];
is_residue = isalpha(msa->aseq[i][pos]);
if (msa->rf) is_consensus = (isalnum(msa->rf[pos]) ? TRUE : FALSE);
else is_consensus = (isalnum(msa->aseq[0][pos]) ? TRUE : FALSE);
if (sym == 'O') sym = 'X';
if (is_consensus) { buf[bpos++] = ( is_residue ? toupper(sym) : '-'); }
else if (is_residue) { buf[bpos++] = tolower(sym); }
else if (! do_dotless) { buf[bpos++] = '.'; }
}
buf[bpos] = '\0';
if (bpos) { if (fprintf(fp, "%s\n", buf) < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "a2m msa file write failed"); }
}
}
} /* end, loop over sequences in the MSA */
free(buf);
return eslOK;
ERROR:
if (buf) free(buf);
return status;
}
/* Function: esl_msafile_psiblast_Write()
* Synopsis: Write an MSA to a stream in PSI-BLAST format
*
* Purpose: Write alignment <msa> in NCBI PSI-BLAST format to
* stream <fp>.
*
* The <msa> should have a valid reference line <msa->rf>,
* with alphanumeric characters marking consensus (match)
* columns, and non-alphanumeric characters marking
* nonconsensus (insert) columns. If it does not have RF
* annotation, then the first sequence in the <msa>
* defines the "consensus".
*
* PSI-BLAST format allows only one symbol ('-') for gaps,
* and cannot represent missing data symbols (Easel's
* '~'). Any missing data symbols are converted to gaps.
*
* Args: fp - open output stream
* msa - MSA to write
*
* Returns: <eslOK> on success.
*
* Throws: <eslEMEM> on allocation failure.
* <eslEWRITE> on any system write failure, such as filled disk.
*/
int
esl_msafile_psiblast_Write(FILE *fp, const ESL_MSA *msa)
{
char *buf = NULL;
int cpl = 60;
int acpl;
int i;
int sym;
int64_t pos, bpos;
int maxnamewidth = esl_str_GetMaxWidth(msa->sqname, msa->nseq);
int is_consensus;
int is_residue;
int status;
ESL_ALLOC(buf, sizeof(char) * (cpl+1));
for (pos = 0; pos < msa->alen; pos += cpl)
{
for (i = 0; i < msa->nseq; i++)
{
acpl = (msa->alen - pos > cpl)? cpl : msa->alen - pos;
#ifdef eslAUGMENT_ALPHABET
if (msa->abc)
{
for (bpos = 0; bpos < acpl; bpos++)
{
sym = msa->abc->sym[msa->ax[i][pos + bpos + 1]];
is_residue = esl_abc_XIsResidue(msa->abc, msa->ax[i][pos+bpos+1]);
if (msa->rf) is_consensus = (isalnum(msa->rf[pos + bpos]) ? TRUE : FALSE);
else is_consensus = (esl_abc_XIsResidue(msa->abc, msa->ax[0][pos+bpos+1]) ? TRUE : FALSE);
if (is_consensus) { buf[bpos] = (is_residue ? toupper(sym) : '-'); }
else { buf[bpos] = (is_residue ? tolower(sym) : '-'); }
}
}
#endif
if (! msa->abc)
{
for (bpos = 0; bpos < acpl; bpos++)
{
sym = msa->aseq[i][pos + bpos];
is_residue = isalnum(sym);
if (msa->rf) is_consensus = (isalnum(msa->rf[pos + bpos]) ? TRUE : FALSE);
else is_consensus = (isalnum(msa->aseq[0][pos+bpos]) ? TRUE : FALSE);
if (is_consensus) { buf[bpos] = (is_residue ? toupper(sym) : '-'); }
else { buf[bpos] = (is_residue ? tolower(sym) : '-'); }
}
}
buf[acpl] = '\0';
if (fprintf(fp, "%-*s %s\n", maxnamewidth, msa->sqname[i], buf) < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "psiblast msa write failed");
} /* end loop over sequences */
if (pos + cpl < msa->alen)
{ if (fputc('\n', fp) < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "psiblast msa write failed"); }
}
free(buf);
return eslOK;
ERROR:
if (buf) free(buf);
return status;
}
/* Function: p7_h2io_WriteASCII()
* Synopsis: Write an H3 HMM in HMMER2 compatible format
*
* Purpose: Write HMM <hmm> to stream <fp> in HMMER2 ASCII save
* file format.
*
* HMMER2 saved the null model and the search configuration
* (local vs. glocal, for example) as part of its HMM file;
* H3 only saves the core HMM. The HMMER2 file is created
* for HMMER2's default ``ls mode'' (glocal) with default
* null model transitions and default special state
* transitions (NECJ).
*
* Optional statistical calibration and alignment checksum
* are not written, because for these H3 and H2 differ too
* much.
*
* Args: fp - stream to write save file format to
* hmm - HMM to save
*
* Returns: <eslOK> on success.
*
* Throws: <eslEMEM> on allocation error.
*
* <eslEINVAL> if <hmm> can't be converted; for example, if
* it is not in a protein or nucleic acid alphabet (H2
* requires biosequence in its save files).
*
* <eslEWRITE> if any write fails; for example, if the
* disk fills up.
*/
int
p7_h2io_WriteASCII(FILE *fp, P7_HMM *hmm)
{
P7_BG *bg; /* H2 saves null model in HMM file */
int k; /* counter for nodes */
int x; /* counter for symbols */
int ts; /* counter for state transitions */
float pmove,ploop; /* default H2 null model transitions */
int status;
if ((bg = p7_bg_Create(hmm->abc)) == NULL) { status = eslEMEM; goto ERROR; }
/* magic header */
if (fprintf(fp, "HMMER2.0 [converted from %s]\n", HMMER_VERSION) < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "h2 profile write failed");
if (fprintf(fp, "NAME %s\n", hmm->name) < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "h2 profile write failed");
if (hmm->acc && fprintf(fp, "ACC %s\n", hmm->acc) < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "h2 profile write failed");
if (hmm->desc && fprintf(fp, "DESC %s\n", hmm->desc) < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "h2 profile write failed");
if (fprintf(fp, "LENG %d\n", hmm->M) < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "h2 profile write failed");
if (hmm->abc->type == eslAMINO) { if (fprintf(fp, "ALPH Amino\n") < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "h2 profile write failed"); }
else if (hmm->abc->type == eslDNA) { if (fprintf(fp, "ALPH Nucleic\n") < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "h2 profile write failed"); }
else if (hmm->abc->type == eslRNA) { if (fprintf(fp, "ALPH Nucleic\n") < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "h2 profile write failed"); }
else ESL_XEXCEPTION(eslEINVAL, "Only protein, DNA, RNA HMMs can be saved in H2 format");
if (fprintf(fp, "RF %s\n", (hmm->flags & p7H_RF) ? "yes" : "no") < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "h2 profile write failed");
if (fprintf(fp, "CS %s\n", (hmm->flags & p7H_CS) ? "yes" : "no") < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "h2 profile write failed");
if (fprintf(fp, "MAP %s\n", (hmm->flags & p7H_MAP) ? "yes" : "no") < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "h2 profile write failed");
/* H3 consensus line has no counterpart in H2 */
if (hmm->comlog != NULL) { if ( (status = h2_multiline(fp, "COM ", hmm->comlog)) != eslOK) goto ERROR; }
if (hmm->nseq != -1) { if ( fprintf (fp, "NSEQ %d\n", hmm->nseq) < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "h2 profile write failed"); }
if (hmm->ctime != NULL) { if ( fprintf (fp, "DATE %s\n", hmm->ctime) < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "h2 profile write failed"); }
/* Checksum is not written; H2 and H3 use different checksum algorithms */
if (hmm->flags & p7H_GA) { if (fprintf(fp, "GA %.1f %.1f\n", hmm->cutoff[p7_GA1], hmm->cutoff[p7_GA2]) < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "h2 profile write failed"); }
if (hmm->flags & p7H_TC) { if (fprintf(fp, "TC %.1f %.1f\n", hmm->cutoff[p7_TC1], hmm->cutoff[p7_TC2]) < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "h2 profile write failed"); }
if (hmm->flags & p7H_NC) { if (fprintf(fp, "NC %.1f %.1f\n", hmm->cutoff[p7_NC1], hmm->cutoff[p7_NC2]) < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "h2 profile write failed"); }
/* in H3, the HMM does not include NECJ; these are part of the profile.
* for emulating H2 output, assume default LS config */
if (fputs("XT ", fp) < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "h2 profile write failed");
pmove = ( (hmm->abc->type == eslAMINO) ? 1./351. : 1./1001.);
ploop = ( (hmm->abc->type == eslAMINO) ? 350./351. : 1000./1001.);
if ( (status = printprob(fp, 6, pmove, 1.0)) != eslOK) goto ERROR; /* NB */
if ( (status = printprob(fp, 6, ploop, 1.0)) != eslOK) goto ERROR; /* NN */
if ( (status = printprob(fp, 6, 0.5, 1.0)) != eslOK) goto ERROR; /* EC */
if ( (status = printprob(fp, 6, 0.5, 1.0)) != eslOK) goto ERROR; /* EJ */
if ( (status = printprob(fp, 6, pmove, 1.0)) != eslOK) goto ERROR; /* CT */
if ( (status = printprob(fp, 6, ploop, 1.0)) != eslOK) goto ERROR; /* CC */
if ( (status = printprob(fp, 6, pmove, 1.0)) != eslOK) goto ERROR; /* JB */
if ( (status = printprob(fp, 6, ploop, 1.0)) != eslOK) goto ERROR; /* JJ */
if (fputc('\n', fp) < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "h2 profile write failed");
/* Save the default H2 null model transitions, not H3's null model transitions */
if (fprintf(fp, "NULT ") < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "h2 profile write failed");
if ( (status = printprob(fp, 6, ploop, 1.0)) != eslOK) goto ERROR; /* 1-p1 */
if ( (status = printprob(fp, 6, pmove, 1.0)) != eslOK) goto ERROR; /* p1 */
if (fputc('\n', fp) < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "h2 profile write failed");
/* but null emissions really are the H3 null model emissions */
if (fputs("NULE ", fp) < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "h2 profile write failed");
for (x = 0; x < hmm->abc->K; x++)
{ if ( (status = printprob(fp, 6, bg->f[x], 1./(float)hmm->abc->K)) != eslOK) goto ERROR; }
if (fputc('\n', fp) < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "h2 profile write failed");
/* Don't save stats; H3 local alignment stats are different from H2 calibration */
/* The main model section */
if (fprintf(fp, "HMM ") < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "h2 profile write failed");
for (x = 0; x < hmm->abc->K; x++)
{ if (fprintf(fp, " %c ", hmm->abc->sym[x]) < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "h2 profile write failed"); }
if (fprintf(fp, "\n") < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "h2 profile write failed");
if (fprintf(fp, " %6s %6s %6s %6s %6s %6s %6s %6s %6s\n",
"m->m", "m->i", "m->d", "i->m", "i->i", "d->m", "d->d", "b->m", "m->e") < 0)
ESL_XEXCEPTION_SYS(eslEWRITE, "h2 profile write failed");
/* Print HMM parameters (main section of the save file) */
if (fprintf(fp, " ") < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "h2 profile write failed");
if ( (status = printprob(fp, 6, 1.-hmm->t[0][p7H_MD], 1.0)) != eslOK) goto ERROR;
if (fprintf(fp, " %6s", "*") < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "h2 profile write failed");
if ( (status = printprob(fp, 6, hmm->t[0][p7H_MD], 1.0)) != eslOK) goto ERROR;
if (fputc('\n', fp) < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "h2 profile write failed");
for (k = 1; k <= hmm->M; k++)
{
/* Line 1: k, match emissions, map */
if (fprintf(fp, " %5d ", k) < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "h2 profile write failed");
for (x = 0; x < hmm->abc->K; x++)
if ( (status = printprob(fp, 6, hmm->mat[k][x], bg->f[x])) != eslOK) goto ERROR;
if (hmm->flags & p7H_MAP)
{ if (fprintf(fp, " %5d", hmm->map[k]) < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "h2 profile write failed"); }
if (fputc('\n', fp) < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "h2 profile write failed");
/* Line 2: RF and insert emissions */
if (fprintf(fp, " %5c ", hmm->flags & p7H_RF ? hmm->rf[k] : '-') < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "h2 profile write failed");
for (x = 0; x < hmm->abc->K; x++)
if ( (status = printprob(fp, 6, ((k < hmm->M) ? hmm->ins[k][x] : 0.0), bg->f[x])) != eslOK) goto ERROR;
if (fputc('\n', fp) < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "h2 profile write failed");
/* Line 3: CS and transition probs */
if (fprintf(fp, " %5c ", hmm->flags & p7H_CS ? hmm->cs[k] : '-') < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "h2 profile write failed");
for (ts = 0; ts < 7; ts++)
if ( (status = printprob(fp, 6, ((k < hmm->M) ? hmm->t[k][ts] : 0.0), 1.0)) != eslOK) goto ERROR;
if ( (status = printprob(fp, 6, ((k==1) ? hmm->t[0][p7H_MM] : 0.0), 1.0)) != eslOK) goto ERROR;
if ( (status = printprob(fp, 6, ((k<hmm->M) ? 0.0: 1.0), 1.0)) != eslOK) goto ERROR;
if (fputc('\n', fp) < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "h2 profile write failed");
}
if (fputs("//\n", fp) < 0) ESL_XEXCEPTION_SYS(eslEWRITE, "h2 profile write failed");
p7_bg_Destroy(bg);
return eslOK;
ERROR:
p7_bg_Destroy(bg);
return status;
}
