/* Function: hmmpgmd_GetRanges()
* Synopsis: Parse command flag into range(s)
*
* Purpose: Given a command flag string <rangestr> of the form
* <start1>..<end1>,<start2>..<end2>...
* parse the string into a RANGE_LIST <list>
*
* Returns: <eslOK> on success <TRUE>, <eslEMEM> on memory allocation failure,
* otherwise <eslESYNTAX> or <eslFAIL> on parsing errors.
*/
int
hmmpgmd_GetRanges (RANGE_LIST *list, char *rangestr) {
char *range;
char *rangestr_cpy;
char *rangestr_cpy_ptr;
int status;
list->N = 0;
list->starts = NULL;
list->ends = NULL;
//first pass to figure out how much to allocate
esl_strdup(rangestr, -1, &rangestr_cpy); // do this because esl_strtok modifies the string, and we shouldn't change the opts value
rangestr_cpy_ptr = rangestr_cpy; // do this because esl_strtok advances the pointer on the target string, but we need to free it
while ( (status = esl_strtok(&rangestr_cpy, ",", &range) ) == eslOK) list->N++;
ESL_ALLOC(list->starts, list->N * sizeof(int));
ESL_ALLOC(list->ends, list->N * sizeof(int));
free(rangestr_cpy_ptr);
//2nd pass to get the values
list->N = 0;
esl_strdup(rangestr, -1, &rangestr_cpy);
rangestr_cpy_ptr = rangestr_cpy;
while ( (status = esl_strtok(&rangestr_cpy, ",", &range) ) == eslOK) {
status = esl_regexp_ParseCoordString(range, list->starts + list->N, list->ends + list->N);
if (status == eslESYNTAX) esl_fatal("--seqdb_ranges takes coords <from>..<to>; %s not recognized", range);
if (status == eslFAIL) esl_fatal("Failed to find <from> or <to> coord in %s", range);
list->N++;
}
free(rangestr_cpy_ptr);
return eslOK;
ERROR:
return eslEMEM;
}
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;
}
int
main(int argc, char **argv)
{
ESL_GETOPTS *go = NULL; /* application configuration */
char *seqfile = NULL; /* sequence file name */
int infmt = eslSQFILE_UNKNOWN; /* format code for seqfile */
ESL_SQFILE *sqfp = NULL; /* open sequence file */
FILE *ofp = NULL; /* output stream for sequences */
int status; /* easel return code */
/***********************************************
* Parse command line
***********************************************/
go = esl_getopts_Create(options);
if (esl_opt_ProcessCmdline(go, argc, argv) != eslOK) cmdline_failure(argv[0], "Failed to parse command line: %s\n", go->errbuf);
if (esl_opt_VerifyConfig(go) != eslOK) cmdline_failure(argv[0], "Error in configuration: %s\n", go->errbuf);
if (esl_opt_GetBoolean(go, "-h") ) cmdline_help (argv[0], go);
if (esl_opt_ArgNumber(go) < 1) cmdline_failure(argv[0], "Incorrect number of command line arguments.\n");
/* Open the sequence file */
seqfile = esl_opt_GetArg(go, 1);
if (esl_opt_GetString(go, "--informat") != NULL) {
infmt = esl_sqio_EncodeFormat(esl_opt_GetString(go, "--informat"));
if (infmt == eslSQFILE_UNKNOWN) esl_fatal("%s is not a valid input sequence file format for --informat", esl_opt_GetString(go, "--informat"));
}
status = esl_sqfile_Open(seqfile, infmt, NULL, &sqfp);
if (status == eslENOTFOUND) cmdline_failure(argv[0], "Sequence file %s not found.\n", seqfile);
else if (status == eslEFORMAT) cmdline_failure(argv[0], "Format of file %s unrecognized.\n", seqfile);
else if (status == eslEINVAL) cmdline_failure(argv[0], "Can't autodetect stdin or .gz.\n");
else if (status != eslOK) cmdline_failure(argv[0], "Open failed, code %d.\n", status);
/* Open the output file, if any */
if (esl_opt_GetBoolean(go, "-O"))
{
if ((ofp = fopen(esl_opt_GetArg(go, 2), "w")) == NULL)
cmdline_failure(argv[0], "Failed to open output file %s\n", esl_opt_GetArg(go, 2));
}
else if (esl_opt_GetString(go, "-o") != NULL)
{
if ((ofp = fopen(esl_opt_GetString(go, "-o"), "w")) == NULL)
cmdline_failure(argv[0], "Failed to open output file %s\n", esl_opt_GetString(go, "-o"));
}
else ofp = stdout;
/* Indexing mode */
if (esl_opt_GetBoolean(go, "--index"))
{
if (esl_opt_ArgNumber(go) != 1) cmdline_failure(argv[0], "Incorrect number of command line arguments.\n");
if (sqfp->data.ascii.do_gzip) cmdline_failure(argv[0], "Can't index a .gz compressed file");
if (sqfp->data.ascii.do_stdin) cmdline_failure(argv[0], "Can't index a standard input pipe");
create_ssi_index(go, sqfp);
}
/* List retrieval mode */
else if (esl_opt_GetBoolean(go, "-f"))
{
if (esl_opt_ArgNumber(go) != 2) cmdline_failure(argv[0], "Incorrect number of command line arguments.\n");
/* Open the SSI index for retrieval */
if (! sqfp->data.ascii.do_gzip && ! sqfp->data.ascii.do_stdin && ! esl_sqio_IsAlignment(sqfp->format))
{
status = esl_sqfile_OpenSSI(sqfp, NULL);
if (status == eslEFORMAT) cmdline_failure(argv[0], "SSI index is in incorrect format\n");
else if (status == eslERANGE) cmdline_failure(argv[0], "SSI index is in 64-bit format and we can't read it\n");
else if (status != eslOK) cmdline_failure(argv[0], "Failed to open SSI index\n");
}
if (esl_opt_GetBoolean(go, "-C")) multifetch_subseq(go, ofp, esl_opt_GetArg(go, 2), sqfp);
else multifetch (go, ofp, esl_opt_GetArg(go, 2), sqfp);
}
/* Single sequence retrieval mode */
else
{
if (esl_opt_ArgNumber(go) != 2) cmdline_failure(argv[0], "Incorrect number of command line arguments.\n");
char *key = esl_opt_GetArg(go, 2);
char *cstring = esl_opt_GetString(go, "-c");
char *newname = esl_opt_GetString(go, "-n");
/* Open the SSI index for retrieval */
if (! sqfp->data.ascii.do_gzip && ! sqfp->data.ascii.do_stdin && ! esl_sqio_IsAlignment(sqfp->format))
{
status = esl_sqfile_OpenSSI(sqfp, NULL);
if (status == eslEFORMAT) cmdline_failure(argv[0], "SSI index is in incorrect format\n");
else if (status == eslERANGE) cmdline_failure(argv[0], "SSI index is in 64-bit format and we can't read it\n");
else if (status != eslOK) cmdline_failure(argv[0], "Failed to open SSI index\n");
}
/* -c: subsequence retrieval; else full sequence retrieval */
if (cstring != NULL)
{
uint32_t start, end;
status = esl_regexp_ParseCoordString(cstring, &start, &end);
if (status == eslESYNTAX) esl_fatal("-c takes arg of subseq coords <from>..<to>; %s not recognized", cstring);
if (status == eslFAIL) esl_fatal("Failed to find <from> or <to> coord in %s", cstring);
onefetch_subseq(go, ofp, sqfp, newname, key, start, end);
if (ofp != stdout) printf("\n\nRetrieved subsequence %s/%d-%d.\n", key, start, end);
}
else
{
onefetch(go, ofp, esl_opt_GetArg(go, 2), sqfp);
if (ofp != stdout) printf("\n\nRetrieved sequence %s.\n", esl_opt_GetArg(go, 2));
}
}
esl_sqfile_Close(sqfp);
esl_getopts_Destroy(go);
return 0;
}
