/*
* Get the next motif. If there is no motif ready then
* returns null.
*/
MOTIF_T* mread_next_motif(MREAD_T *mread) {
if (mread_has_motif(mread)) {
return post_process_motif(mread,
mread->formats->next_motif(mread->formats->data));
}
return NULL;
}
/*************************************************************************
* Read a motif database
*************************************************************************/
static MOTIF_DB_T* read_motifs(int id, char* motif_source, char* bg_source,
ARRAY_T** bg, double pseudocount, RBTREE_T *selected, ALPH_T alph) {
// vars
int read_motifs;
MOTIF_DB_T* motifdb;
MREAD_T *mread;
MOTIF_T *motif;
ARRAYLST_T *motifs;
// open the motif file for reading
mread = mread_create(motif_source, OPEN_MFILE);
mread_set_pseudocount(mread, pseudocount);
// determine background to use
if (*bg != NULL) mread_set_background(mread, *bg);
else mread_set_bg_source(mread, bg_source);
// load motifs
read_motifs = 0;
if (rbtree_size(selected) > 0) {
motifs = arraylst_create();
while(mread_has_motif(mread)) {
motif = mread_next_motif(mread);
read_motifs++;
if (rbtree_find(selected, get_motif_id(motif))) {
arraylst_add(motif, motifs);
} else {
DEBUG_FMT(NORMAL_VERBOSE, "Discarding motif %s in %s.\n",
get_motif_id(motif), motif_source);
destroy_motif(motif);
}
}
} else {
motifs = mread_load(mread, NULL);
read_motifs = arraylst_size(motifs);
}
arraylst_fit(motifs);
if (read_motifs > 0) {
// check the alphabet
if (mread_get_alphabet(mread) != alph) {
die("Expected %s alphabet motifs\n", alph_name(alph));
}
// get the background
if (*bg == NULL) *bg = mread_get_background(mread);
} else {
fprintf(stderr, "Warning: Motif file %s contains no motifs.\n", motif_source);
}
// clean up motif reader
mread_destroy(mread);
// create motif db
motifdb = mm_malloc(sizeof(MOTIF_DB_T));
memset(motifdb, 0, sizeof(MOTIF_DB_T));
motifdb->id = id;
motifdb->source = strdup(motif_source);
motifdb->motifs = motifs;
return motifdb;
}
/**************************************************************************
* Generate logos for all motifs in a file
**************************************************************************/
static void generate_file_logos(OPTIONS_T *options) {
STR_T *path;
MREAD_T *mread;
MOTIF_T *motif;
// file path buffer
path = str_create(100);
str_append(path, options->dir, strlen(options->dir));
if (str_char(path, -1) != '/') str_append(path, "/", 1);
// create output directory
if (create_output_directory(str_internal(path), TRUE, FALSE))
exit(EXIT_FAILURE);
// open motif file
mread = mread_create(options->motifs_file, OPEN_MFILE);
while (mread_has_motif(mread)) {
motif = mread_next_motif(mread);
generate_motif_logos(options, path, motif);
destroy_motif(motif);
}
mread_destroy(mread);
str_destroy(path, FALSE);
}
ARRAYLST_T* load_motifs(AMA_OPTIONS_T *opts) {
ARRAYLST_T *motifs;
ARRAY_T *pos_bg_freqs, *rev_bg_freqs;
MREAD_T *mread;
MOTIF_T *motif, *motif_rc;
double range;
PSSM_T *pos_pssm, *neg_pssm;
int total_motifs;
ALPH_T *alph;
//
// Read the motifs and background model.
//
//this reads any meme file, xml, txt and html
mread = mread_create(opts->motif_filename, OPEN_MFILE);
mread_set_bg_source(mread, opts->bg_filename);
mread_set_pseudocount(mread, opts->pseudocount);
// sanity check, since the rest of the code relies on the motifs being complementable
alph = alph_hold(mread_get_alphabet(mread));
if (alph == NULL) die("Unable to determine alphabet from motifs");
if (opts->scan_both_strands && !alph_has_complement(alph)) {
opts->scan_both_strands = false;
}
if (opts->num_gc_bins > 1 && alph_size_core(alph) != 4 && alph_size_pairs(alph) != 2) {
fprintf(stderr, "Warning: The motif alphabet does not have exactly 2 complementary pairs so \"GC binning\" will be disabled.\n");
opts->num_gc_bins = 1;
}
pos_bg_freqs = mread_get_background(mread);
rev_bg_freqs = NULL;
if (opts->scan_both_strands) {
rev_bg_freqs = allocate_array(get_array_length(pos_bg_freqs));
copy_array(pos_bg_freqs, rev_bg_freqs);
complement_swap_freqs(alph, rev_bg_freqs, rev_bg_freqs);
}
// allocate memory for motifs
motifs = arraylst_create();
//
// Convert motif matrices into log-odds matrices.
// Scale them.
// Compute the lookup tables for the PDF of scaled log-odds scores.
//
range = 300; // 100 is not very good; 1000 is great but too slow
neg_pssm = NULL;
total_motifs = 0;
while (mread_has_motif(mread)) {
motif = mread_next_motif(mread);
total_motifs++;
if (rbtree_size(opts->selected_motifs) == 0 || rbtree_find(opts->selected_motifs, get_motif_id(motif)) != NULL) {
if (verbosity >= HIGH_VERBOSE) {
fprintf(stderr, "Using motif %s of width %d.\n", get_motif_id(motif), get_motif_length(motif));
}
pos_pssm =
build_motif_pssm(
motif,
pos_bg_freqs,
pos_bg_freqs,
NULL, // Priors not used
0.0L, // alpha not used
range,
opts->num_gc_bins,
true
);
//
// Note: If scanning both strands, we complement the motif frequencies
// but not the background frequencies so the motif looks the same.
// However, the given frequencies are used in computing the p-values
// since they represent the frequencies on the negative strands.
// (If we instead were to complement the input sequence, keeping the
// the motif fixed, we would need to use the complemented frequencies
// in computing the p-values. Is that any clearer?)
//
if (opts->scan_both_strands) {
motif_rc = dup_rc_motif(motif);
neg_pssm =
build_motif_pssm(
motif_rc,
rev_bg_freqs,
pos_bg_freqs,
NULL, // Priors not used
0.0L, // alpha not used
range,
opts->num_gc_bins,
true
);
destroy_motif(motif_rc);
}
arraylst_add(motif_and_pssm_create(motif, pos_pssm, neg_pssm), motifs);
} else {
if (verbosity >= HIGH_VERBOSE) fprintf(stderr, "Skipping motif %s.\n",
get_motif_id(motif));
destroy_motif(motif);
}
}
mread_destroy(mread);
free_array(pos_bg_freqs);
free_array(rev_bg_freqs);
alph_release(alph);
if (verbosity >= NORMAL_VERBOSE) {
fprintf(stderr, "Loaded %d/%d motifs from %s.\n",
arraylst_size(motifs), total_motifs, opts->motif_filename);
}
return motifs;
}