redesigned indexing and option APIs

This commit is contained in:
Heng Li
2017-09-14 17:02:01 -04:00
parent 0f7455cefa
commit eb00521d9b
5 changed files with 151 additions and 98 deletions
+23 -80
View File
@@ -25,7 +25,7 @@ void liftrlimit() {}
static struct option long_options[] = {
{ "bucket-bits", required_argument, 0, 0 },
{ "mb-size", required_argument, 0, 'K' },
{ "int-rname", no_argument, 0, 0 },
{ "int-rname", no_argument, 0, 0 }, // obsolete; kept as a placeholder
{ "no-kalloc", no_argument, 0, 0 },
{ "print-qname", no_argument, 0, 0 },
{ "no-self", no_argument, 0, 0 },
@@ -62,21 +62,21 @@ static inline int64_t mm_parse_num(const char *str)
int main(int argc, char *argv[])
{
mm_mapopt_t opt;
int i, c, k = 15, w = -1, bucket_bits = MM_IDX_DEF_B, n_threads = 3, keep_name = 1, is_idx, is_hpc = 0, long_idx, idx_par_set = 0, max_intron_len = 0, n_idx_part = 0;
int minibatch_size = 200000000;
uint64_t batch_size = 4000000000ULL;
mm_bseq_file_t *fp = 0;
mm_idxopt_t ipt;
int i, c, n_threads = 3, long_idx, max_intron_len = 0;
char *fnw = 0, *rg = 0, *s;
FILE *fpr = 0, *fpw = 0, *fp_help = stderr;
FILE *fpw = 0, *fp_help = stderr;
mm_idx_reader_t *idx_rdr;
liftrlimit();
mm_realtime0 = realtime();
mm_mapopt_init(&opt);
mm_idxopt_init(&ipt);
while ((c = getopt_long(argc, argv, "aSw:k:K:t:r:f:Vv:g:G:I:d:XT:s:x:Hcp:M:n:z:A:B:O:E:m:N:Qu:R:h", long_options, &long_idx)) >= 0) {
if (c == 'w') w = atoi(optarg), idx_par_set = 1;
else if (c == 'k') k = atoi(optarg), idx_par_set = 1;
else if (c == 'H') is_hpc = 1, idx_par_set = 1;
if (c == 'w') ipt.w = atoi(optarg);
else if (c == 'k') ipt.k = atoi(optarg);
else if (c == 'H') ipt.is_hpc = 1;
else if (c == 'd') fnw = optarg; // the above are indexing related options, except -I
else if (c == 'r') opt.bw = (int)mm_parse_num(optarg);
else if (c == 't') n_threads = atoi(optarg);
@@ -98,12 +98,11 @@ int main(int argc, char *argv[])
else if (c == 'B') opt.b = atoi(optarg);
else if (c == 'z') opt.zdrop = atoi(optarg);
else if (c == 's') opt.min_dp_max = atoi(optarg);
else if (c == 'I') batch_size = mm_parse_num(optarg);
else if (c == 'K') minibatch_size = (int)mm_parse_num(optarg);
else if (c == 'I') ipt.batch_size = mm_parse_num(optarg);
else if (c == 'K') ipt.mini_batch_size = (int)mm_parse_num(optarg);
else if (c == 'R') rg = optarg;
else if (c == 'h') fp_help = stdout;
else if (c == 0 && long_idx == 0) bucket_bits = atoi(optarg); // --bucket-bits
else if (c == 0 && long_idx == 2) keep_name = 0; // --int-rname
else if (c == 0 && long_idx == 0) ipt.bucket_bits = atoi(optarg); // --bucket-bits
else if (c == 0 && long_idx == 3) mm_dbg_flag |= MM_DBG_NO_KALLOC; // --no-kalloc
else if (c == 0 && long_idx == 4) mm_dbg_flag |= MM_DBG_PRINT_QNAME; // --print-qname
else if (c == 0 && long_idx == 5) opt.flag |= MM_F_NO_SELF; // --no-self
@@ -140,55 +139,12 @@ int main(int argc, char *argv[])
opt.e = opt.e2 = strtol(optarg, &s, 10);
if (*s == ',') opt.e2 = strtol(s + 1, &s, 10);
} else if (c == 'x') {
if (strcmp(optarg, "ava-ont") == 0) {
opt.flag |= MM_F_AVA | MM_F_NO_SELF;
opt.min_chain_score = 100, opt.pri_ratio = 0.0f, opt.max_gap = 10000, opt.max_chain_skip = 25;
minibatch_size = 500000000;
k = 15, w = 5;
} else if (strcmp(optarg, "ava-pb") == 0) {
opt.flag |= MM_F_AVA | MM_F_NO_SELF;
opt.min_chain_score = 100, opt.pri_ratio = 0.0f, opt.max_gap = 10000, opt.max_chain_skip = 25;
minibatch_size = 500000000;
is_hpc = 1, k = 19, w = 5;
} else if (strcmp(optarg, "map10k") == 0 || strcmp(optarg, "map-pb") == 0) {
is_hpc = 1, k = 19;
} else if (strcmp(optarg, "map-ont") == 0) {
is_hpc = 0, k = 15;
} else if (strcmp(optarg, "asm5") == 0) {
k = 19, w = 19;
opt.a = 1, opt.b = 19, opt.q = 39, opt.q2 = 81, opt.e = 3, opt.e2 = 1, opt.zdrop = 200;
opt.min_dp_max = 200;
} else if (strcmp(optarg, "asm10") == 0) {
k = 19, w = 19;
opt.a = 1, opt.b = 9, opt.q = 16, opt.q2 = 41, opt.e = 2, opt.e2 = 1, opt.zdrop = 200;
opt.min_dp_max = 200;
} else if (strcmp(optarg, "short") == 0 || strcmp(optarg, "sr") == 0) {
k = 21, w = 11, is_hpc = 0;
minibatch_size = 50000000;
opt.flag |= MM_F_APPROX_EXT;
opt.a = 2, opt.b = 8, opt.q = 12, opt.e = 2, opt.q2 = 32, opt.e2 = 1;
opt.max_gap = 100;
opt.pri_ratio = 0.5f;
opt.min_cnt = 2;
opt.min_chain_score = 20;
opt.min_dp_max = 40;
opt.best_n = 20;
opt.bw = 50;
opt.mid_occ = 1000;
} else if (strcmp(optarg, "splice") == 0 || strcmp(optarg, "cdna") == 0) {
k = 15, w = 5;
opt.flag |= MM_F_SPLICE | MM_F_SPLICE_FOR | MM_F_SPLICE_REV;
opt.max_gap = 2000, opt.max_gap_ref = opt.bw = 200000;
opt.a = 1, opt.b = 2, opt.q = 2, opt.e = 1, opt.q2 = 32, opt.e2 = 0;
opt.noncan = 5;
opt.zdrop = 200;
} else {
if (mm_preset(optarg, &ipt, &opt) < 0) {
fprintf(stderr, "[E::%s] unknown preset '%s'\n", __func__, optarg);
return 1;
}
}
}
if (w < 0) w = (int)(.6666667 * k + .499);
if ((opt.flag & MM_F_SPLICE) && max_intron_len > 0)
opt.max_gap_ref = opt.bw = max_intron_len;
@@ -197,8 +153,8 @@ int main(int argc, char *argv[])
fprintf(fp_help, "Options:\n");
fprintf(fp_help, " Indexing:\n");
fprintf(fp_help, " -H use homopolymer-compressed k-mer\n");
fprintf(fp_help, " -k INT k-mer size (no larger than 28) [%d]\n", k);
fprintf(fp_help, " -w INT minizer window size [{-k}*2/3]\n");
fprintf(fp_help, " -k INT k-mer size (no larger than 28) [%d]\n", ipt.k);
fprintf(fp_help, " -w INT minizer window size [%d]\n", ipt.w);
fprintf(fp_help, " -I NUM split index for every ~NUM input bases [4G]\n");
fprintf(fp_help, " -d FILE dump index to FILE []\n");
fprintf(fp_help, " Mapping:\n");
@@ -227,7 +183,7 @@ int main(int argc, char *argv[])
fprintf(fp_help, " -c output CIGAR in PAF\n");
fprintf(fp_help, " -S output the cs tag in PAF (cs encodes both query and ref sequences)\n");
fprintf(fp_help, " -t INT number of threads [%d]\n", n_threads);
fprintf(fp_help, " -K NUM minibatch size [200M]\n");
fprintf(fp_help, " -K NUM minibatch size for mapping [200M]\n");
// fprintf(fp_help, " -v INT verbose level [%d]\n", mm_verbose);
fprintf(fp_help, " --version show version number\n");
fprintf(fp_help, " Preset:\n");
@@ -244,33 +200,22 @@ int main(int argc, char *argv[])
return fp_help == stdout? 0 : 1;
}
is_idx = mm_idx_is_idx(argv[optind]);
if (is_idx < 0) {
idx_rdr = mm_idx_reader_open(argv[optind], &ipt);
if (idx_rdr == 0) {
fprintf(stderr, "[ERROR] failed to open file '%s'\n", argv[optind]);
return 1;
}
if (!is_idx && fnw == 0 && argc - optind < 2) {
if (!idx_rdr->is_idx && fnw == 0 && argc - optind < 2) {
fprintf(stderr, "[ERROR] missing input: please specify a query file to map or option -d to keep the index\n");
return 1;
}
if (is_idx) fpr = fopen(argv[optind], "rb");
else fp = mm_bseq_open(argv[optind]);
if (fnw) fpw = fopen(fnw, "wb");
if (opt.flag & MM_F_OUT_SAM)
mm_write_sam_hdr_no_SQ(rg, MM_VERSION, argc, argv);
for (;;) {
mm_idx_t *mi;
if (fpr) {
mi = mm_idx_load(fpr);
if (mi == 0) break;
if (idx_par_set && mm_verbose >= 2 && (mi->k != k || mi->w != w || mi->is_hpc != is_hpc))
fprintf(stderr, "[WARNING] \033[1;31mIndexing parameters on the command line (-k/-w/-H) overridden by parameters in the prebuilt index.\033[0m\n");
} else {
mi = mm_idx_gen(fp, w, k, bucket_bits, is_hpc, minibatch_size, n_threads, batch_size, keep_name);
}
if (mi == 0) break;
++n_idx_part;
if (mm_verbose >= 2 && n_idx_part > 1 && (opt.flag&MM_F_OUT_SAM) && !(opt.flag&MM_F_NO_SAM_SQ))
if ((mi = mm_idx_reader_read(idx_rdr, n_threads)) == 0) break;
if (mm_verbose >= 2 && idx_rdr->n_parts > 1 && (opt.flag&MM_F_OUT_SAM) && !(opt.flag&MM_F_NO_SAM_SQ))
fprintf(stderr, "[WARNING] \033[1;31mSAM output is malformated due to internal @SQ lines. Please add option --no-sam-sq or filter afterwards.\033[0m\n");
if (mm_verbose >= 3)
fprintf(stderr, "[M::%s::%.3f*%.2f] loaded/built the index for %d target sequence(s)\n",
@@ -278,17 +223,15 @@ int main(int argc, char *argv[])
if (fpw) {
mm_idx_dump(fpw, mi);
if (mm_verbose >= 3)
fprintf(stderr, "[M::%s::%.3f*%.2f] dumpped the (partial) index to disk\n", __func__, realtime() - mm_realtime0, cputime() / (realtime() - mm_realtime0));
fprintf(stderr, "[M::%s::%.3f*%.2f] dumpped the current part of the index to disk\n", __func__, realtime() - mm_realtime0, cputime() / (realtime() - mm_realtime0));
}
if (argc != optind + 1) mm_mapopt_update(&opt, mi);
if (mm_verbose >= 3) mm_idx_stat(mi);
for (i = optind + 1; i < argc; ++i)
mm_map_file(mi, argv[i], &opt, n_threads, minibatch_size);
mm_map_file(mi, argv[i], &opt, n_threads);
mm_idx_destroy(mi);
}
if (fpw) fclose(fpw);
if (fpr) fclose(fpr);
if (fp) mm_bseq_close(fp);
fprintf(stderr, "[M::%s] Version: %s\n", __func__, MM_VERSION);
fprintf(stderr, "[M::%s] CMD:", __func__);