Mercurial > repos > siyuan > prada

comparison pyPRADA_1.2/tools/samtools-0.1.16/sam_view.c @ 0:acc2ca1a3ba4

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
Uploaded
author siyuan
date Thu, 20 Feb 2014 00:44:58 -0500
parents
children
comparison
equal deleted inserted replaced
-1:000000000000 0:acc2ca1a3ba4
1 #include <stdlib.h>
2 #include <string.h>
3 #include <stdio.h>
4 #include <unistd.h>
5 #include <math.h>
6 #include "sam_header.h"
7 #include "sam.h"
8 #include "faidx.h"
9 #include "kstring.h"
10 #include "khash.h"
11 KHASH_SET_INIT_STR(rg)
12
13 // When counting records instead of printing them,
14 // data passed to the bam_fetch callback is encapsulated in this struct.
15 typedef struct {
16 bam_header_t *header;
17 int *count;
18 } count_func_data_t;
19
20 typedef khash_t(rg) *rghash_t;
21
22 static rghash_t g_rghash = 0;
23 static int g_min_mapQ = 0, g_flag_on = 0, g_flag_off = 0;
24 static char *g_library, *g_rg;
25 static int g_sol2sanger_tbl[128];
26 static void *g_bed;
27
28 void *bed_read(const char *fn);
29 void bed_destroy(void *_h);
30 int bed_overlap(const void *_h, const char *chr, int beg, int end);
31
32 static void sol2sanger(bam1_t *b)
33 {
34 int l;
35 uint8_t *qual = bam1_qual(b);
36 if (g_sol2sanger_tbl[30] == 0) {
37 for (l = 0; l != 128; ++l) {
38 g_sol2sanger_tbl[l] = (int)(10.0 * log(1.0 + pow(10.0, (l - 64 + 33) / 10.0)) / log(10.0) + .499);
39 if (g_sol2sanger_tbl[l] >= 93) g_sol2sanger_tbl[l] = 93;
40 }
41 }
42 for (l = 0; l < b->core.l_qseq; ++l) {
43 int q = qual[l];
44 if (q > 127) q = 127;
45 qual[l] = g_sol2sanger_tbl[q];
46 }
47 }
48
49 static inline int __g_skip_aln(const bam_header_t *h, const bam1_t *b)
50 {
51 if (b->core.qual < g_min_mapQ || ((b->core.flag & g_flag_on) != g_flag_on) || (b->core.flag & g_flag_off))
52 return 1;
53 if (g_bed && b->core.tid >= 0 && !bed_overlap(g_bed, h->target_name[b->core.tid], b->core.pos, bam_calend(&b->core, bam1_cigar(b))))
54 return 1;
55 if (g_rg || g_rghash) {
56 uint8_t *s = bam_aux_get(b, "RG");
57 if (s) {
58 if (g_rg) return (strcmp(g_rg, (char*)(s + 1)) == 0)? 0 : 1;
59 if (g_rghash) {
60 khint_t k = kh_get(rg, g_rghash, (char*)(s + 1));
61 return (k != kh_end(g_rghash))? 0 : 1;
62 }
63 }
64 }
65 if (g_library) {
66 const char *p = bam_get_library((bam_header_t*)h, b);
67 return (p && strcmp(p, g_library) == 0)? 0 : 1;
68 }
69 return 0;
70 }
71
72 static char *drop_rg(char *hdtxt, rghash_t h, int *len)
73 {
74 char *p = hdtxt, *q, *r, *s;
75 kstring_t str;
76 memset(&str, 0, sizeof(kstring_t));
77 while (1) {
78 int toprint = 0;
79 q = strchr(p, '\n');
80 if (q == 0) q = p + strlen(p);
81 if (q - p < 3) break; // the line is too short; then stop
82 if (strncmp(p, "@RG\t", 4) == 0) {
83 int c;
84 khint_t k;
85 if ((r = strstr(p, "\tID:")) != 0) {
86 r += 4;
87 for (s = r; *s != '\0' && *s != '\n' && *s != '\t'; ++s);
88 c = *s; *s = '\0';
89 k = kh_get(rg, h, r);
90 *s = c;
91 if (k != kh_end(h)) toprint = 1;
92 }
93 } else toprint = 1;
94 if (toprint) {
95 kputsn(p, q - p, &str); kputc('\n', &str);
96 }
97 p = q + 1;
98 }
99 *len = str.l;
100 return str.s;
101 }
102
103 // callback function for bam_fetch() that prints nonskipped records
104 static int view_func(const bam1_t *b, void *data)
105 {
106 if (!__g_skip_aln(((samfile_t*)data)->header, b))
107 samwrite((samfile_t*)data, b);
108 return 0;
109 }
110
111 // callback function for bam_fetch() that counts nonskipped records
112 static int count_func(const bam1_t *b, void *data)
113 {
114 if (!__g_skip_aln(((count_func_data_t*)data)->header, b)) {
115 (*((count_func_data_t*)data)->count)++;
116 }
117 return 0;
118 }
119
120 static int usage(int is_long_help);
121
122 int main_samview(int argc, char *argv[])
123 {
124 int c, is_header = 0, is_header_only = 0, is_bamin = 1, ret = 0, compress_level = -1, is_bamout = 0, slx2sngr = 0, is_count = 0;
125 int of_type = BAM_OFDEC, is_long_help = 0;
126 int count = 0;
127 samfile_t *in = 0, *out = 0;
128 char in_mode[5], out_mode[5], *fn_out = 0, *fn_list = 0, *fn_ref = 0, *fn_rg = 0;
129
130 /* parse command-line options */
131 strcpy(in_mode, "r"); strcpy(out_mode, "w");
132 while ((c = getopt(argc, argv, "Sbct:h1Ho:q:f:F:ul:r:xX?T:CR:L:")) >= 0) {
133 switch (c) {
134 case 'c': is_count = 1; break;
135 case 'C': slx2sngr = 1; break;
136 case 'S': is_bamin = 0; break;
137 case 'b': is_bamout = 1; break;
138 case 't': fn_list = strdup(optarg); is_bamin = 0; break;
139 case 'h': is_header = 1; break;
140 case 'H': is_header_only = 1; break;
141 case 'o': fn_out = strdup(optarg); break;
142 case 'f': g_flag_on = strtol(optarg, 0, 0); break;
143 case 'F': g_flag_off = strtol(optarg, 0, 0); break;
144 case 'q': g_min_mapQ = atoi(optarg); break;
145 case 'u': compress_level = 0; break;
146 case '1': compress_level = 1; break;
147 case 'l': g_library = strdup(optarg); break;
148 case 'L': g_bed = bed_read(optarg); break;
149 case 'r': g_rg = strdup(optarg); break;
150 case 'R': fn_rg = strdup(optarg); break;
151 case 'x': of_type = BAM_OFHEX; break;
152 case 'X': of_type = BAM_OFSTR; break;
153 case '?': is_long_help = 1; break;
154 case 'T': fn_ref = strdup(optarg); is_bamin = 0; break;
155 default: return usage(is_long_help);
156 }
157 }
158 if (compress_level >= 0) is_bamout = 1;
159 if (is_header_only) is_header = 1;
160 if (is_bamout) strcat(out_mode, "b");
161 else {
162 if (of_type == BAM_OFHEX) strcat(out_mode, "x");
163 else if (of_type == BAM_OFSTR) strcat(out_mode, "X");
164 }
165 if (is_bamin) strcat(in_mode, "b");
166 if (is_header) strcat(out_mode, "h");
167 if (compress_level >= 0) {
168 char tmp[2];
169 tmp[0] = compress_level + '0'; tmp[1] = '\0';
170 strcat(out_mode, tmp);
171 }
172 if (argc == optind) return usage(is_long_help); // potential memory leak...
173
174 // read the list of read groups
175 if (fn_rg) {
176 FILE *fp_rg;
177 char buf[1024];
178 int ret;
179 g_rghash = kh_init(rg);
180 fp_rg = fopen(fn_rg, "r");
181 while (!feof(fp_rg) && fscanf(fp_rg, "%s", buf) > 0) // this is not a good style, but bear me...
182 kh_put(rg, g_rghash, strdup(buf), &ret); // we'd better check duplicates...
183 fclose(fp_rg);
184 }
185
186 // generate the fn_list if necessary
187 if (fn_list == 0 && fn_ref) fn_list = samfaipath(fn_ref);
188 // open file handlers
189 if ((in = samopen(argv[optind], in_mode, fn_list)) == 0) {
190 fprintf(stderr, "[main_samview] fail to open \"%s\" for reading.\n", argv[optind]);
191 ret = 1;
192 goto view_end;
193 }
194 if (in->header == 0) {
195 fprintf(stderr, "[main_samview] fail to read the header from \"%s\".\n", argv[optind]);
196 ret = 1;
197 goto view_end;
198 }
199 if (g_rghash) { // FIXME: I do not know what "bam_header_t::n_text" is for...
200 char *tmp;
201 int l;
202 tmp = drop_rg(in->header->text, g_rghash, &l);
203 free(in->header->text);
204 in->header->text = tmp;
205 in->header->l_text = l;
206 }
207 if (!is_count && (out = samopen(fn_out? fn_out : "-", out_mode, in->header)) == 0) {
208 fprintf(stderr, "[main_samview] fail to open \"%s\" for writing.\n", fn_out? fn_out : "standard output");
209 ret = 1;
210 goto view_end;
211 }
212 if (is_header_only) goto view_end; // no need to print alignments
213
214 if (argc == optind + 1) { // convert/print the entire file
215 bam1_t *b = bam_init1();
216 int r;
217 while ((r = samread(in, b)) >= 0) { // read one alignment from `in'
218 if (!__g_skip_aln(in->header, b)) {
219 if (slx2sngr) sol2sanger(b);
220 if (!is_count) samwrite(out, b); // write the alignment to `out'
221 count++;
222 }
223 }
224 if (r < -1) {
225 fprintf(stderr, "[main_samview] truncated file.\n");
226 ret = 1;
227 }
228 bam_destroy1(b);
229 } else { // retrieve alignments in specified regions
230 int i;
231 bam_index_t *idx = 0;
232 if (is_bamin) idx = bam_index_load(argv[optind]); // load BAM index
233 if (idx == 0) { // index is unavailable
234 fprintf(stderr, "[main_samview] random alignment retrieval only works for indexed BAM files.\n");
235 ret = 1;
236 goto view_end;
237 }
238 for (i = optind + 1; i < argc; ++i) {
239 int tid, beg, end, result;
240 bam_parse_region(in->header, argv[i], &tid, &beg, &end); // parse a region in the format like `chr2:100-200'
241 if (tid < 0) { // reference name is not found
242 fprintf(stderr, "[main_samview] region \"%s\" specifies an unknown reference name. Continue anyway.\n", argv[i]);
243 continue;
244 }
245 // fetch alignments
246 if (is_count) {
247 count_func_data_t count_data = { in->header, &count };
248 result = bam_fetch(in->x.bam, idx, tid, beg, end, &count_data, count_func);
249 } else
250 result = bam_fetch(in->x.bam, idx, tid, beg, end, out, view_func);
251 if (result < 0) {
252 fprintf(stderr, "[main_samview] retrieval of region \"%s\" failed due to truncated file or corrupt BAM index file\n", argv[i]);
253 ret = 1;
254 break;
255 }
256 }
257 bam_index_destroy(idx); // destroy the BAM index
258 }
259
260 view_end:
261 if (is_count && ret == 0) {
262 printf("%d\n", count);
263 }
264 // close files, free and return
265 free(fn_list); free(fn_ref); free(fn_out); free(g_library); free(g_rg); free(fn_rg);
266 if (g_bed) bed_destroy(g_bed);
267 if (g_rghash) {
268 khint_t k;
269 for (k = 0; k < kh_end(g_rghash); ++k)
270 if (kh_exist(g_rghash, k)) free((char*)kh_key(g_rghash, k));
271 kh_destroy(rg, g_rghash);
272 }
273 samclose(in);
274 if (!is_count)
275 samclose(out);
276 return ret;
277 }
278
279 static int usage(int is_long_help)
280 {
281 fprintf(stderr, "\n");
282 fprintf(stderr, "Usage: samtools view [options] <in.bam>|<in.sam> [region1 [...]]\n\n");
283 fprintf(stderr, "Options: -b output BAM\n");
284 fprintf(stderr, " -h print header for the SAM output\n");
285 fprintf(stderr, " -H print header only (no alignments)\n");
286 fprintf(stderr, " -S input is SAM\n");
287 fprintf(stderr, " -u uncompressed BAM output (force -b)\n");
288 fprintf(stderr, " -1 fast compression (force -b)\n");
289 fprintf(stderr, " -x output FLAG in HEX (samtools-C specific)\n");
290 fprintf(stderr, " -X output FLAG in string (samtools-C specific)\n");
291 fprintf(stderr, " -c print only the count of matching records\n");
292 fprintf(stderr, " -L FILE output alignments overlapping the input BED FILE [null]\n");
293 fprintf(stderr, " -t FILE list of reference names and lengths (force -S) [null]\n");
294 fprintf(stderr, " -T FILE reference sequence file (force -S) [null]\n");
295 fprintf(stderr, " -o FILE output file name [stdout]\n");
296 fprintf(stderr, " -R FILE list of read groups to be outputted [null]\n");
297 fprintf(stderr, " -f INT required flag, 0 for unset [0]\n");
298 fprintf(stderr, " -F INT filtering flag, 0 for unset [0]\n");
299 fprintf(stderr, " -q INT minimum mapping quality [0]\n");
300 fprintf(stderr, " -l STR only output reads in library STR [null]\n");
301 fprintf(stderr, " -r STR only output reads in read group STR [null]\n");
302 fprintf(stderr, " -? longer help\n");
303 fprintf(stderr, "\n");
304 if (is_long_help)
305 fprintf(stderr, "Notes:\n\
306 \n\
307 1. By default, this command assumes the file on the command line is in\n\
308 the BAM format and it prints the alignments in SAM. If `-t' is\n\
309 applied, the input file is assumed to be in the SAM format. The\n\
310 file supplied with `-t' is SPACE/TAB delimited with the first two\n\
311 fields of each line consisting of the reference name and the\n\
312 corresponding sequence length. The `.fai' file generated by `faidx'\n\
313 can be used here. This file may be empty if reads are unaligned.\n\
314 \n\
315 2. SAM->BAM conversion: `samtools view -bT ref.fa in.sam.gz'.\n\
316 \n\
317 3. BAM->SAM conversion: `samtools view in.bam'.\n\
318 \n\
319 4. A region should be presented in one of the following formats:\n\
320 `chr1', `chr2:1,000' and `chr3:1000-2,000'. When a region is\n\
321 specified, the input alignment file must be an indexed BAM file.\n\
322 \n\
323 5. Option `-u' is preferred over `-b' when the output is piped to\n\
324 another samtools command.\n\
325 \n\
326 6. In a string FLAG, each character represents one bit with\n\
327 p=0x1 (paired), P=0x2 (properly paired), u=0x4 (unmapped),\n\
328 U=0x8 (mate unmapped), r=0x10 (reverse), R=0x20 (mate reverse)\n\
329 1=0x40 (first), 2=0x80 (second), s=0x100 (not primary), \n\
330 f=0x200 (failure) and d=0x400 (duplicate). Note that `-x' and\n\
331 `-X' are samtools-C specific. Picard and older samtools do not\n\
332 support HEX or string flags.\n\
333 \n");
334 return 1;
335 }
336
337 int main_import(int argc, char *argv[])
338 {
339 int argc2, ret;
340 char **argv2;
341 if (argc != 4) {
342 fprintf(stderr, "Usage: bamtk import <in.ref_list> <in.sam> <out.bam>\n");
343 return 1;
344 }
345 argc2 = 6;
346 argv2 = calloc(6, sizeof(char*));
347 argv2[0] = "import", argv2[1] = "-o", argv2[2] = argv[3], argv2[3] = "-bt", argv2[4] = argv[1], argv2[5] = argv[2];
348 ret = main_samview(argc2, argv2);
349 free(argv2);
350 return ret;
351 }