0
|
1 #!/usr/bin/perl
|
|
2 use strict;
|
|
3 use Getopt::Long;
|
|
4
|
|
5 #
|
|
6 # Filter a pileup file on forward/reverse presence and %read having the variant
|
|
7 # The error code
|
|
8 # 1 : multiple variant type detected insertion/deletion/mutation
|
|
9 # 1i : inconsistency in insertion
|
|
10 # 1d : inconsistency in deletion
|
|
11 # 1m : inconsistency in mutation
|
|
12 # 2 : insufficient depth
|
|
13 # 3 : insufficient variant frequency
|
|
14 # 4 : variant position not covered by forward and reverse reads
|
|
15 # 5 : variant with other variant in neighbourhood
|
|
16 # 6 : too much depth
|
|
17 # 8 : parsing error (couldn't parse the mpileup line correctly)
|
|
18 # 9 : parsing error (couldn't parse the readbase string correctly)
|
|
19
|
|
20
|
|
21 my $inputfile;
|
|
22 my $logfile;
|
|
23 my $MIN_DISTANCE=0;
|
|
24 my $MIN_VARIANTFREQUENCY=0;
|
|
25 my $MIN_FORWARDREVERSE=0;
|
|
26 my $MIN_DEPTH=0;
|
|
27 my $MAX_DEPTH=500;
|
|
28 my $VERBOSE=0;
|
|
29 my $ONLY_UNFILTERED_VARIANT="OFF";
|
|
30 my $DO_STAT="NO";
|
|
31
|
|
32
|
|
33 my $STAT_MIN_DEPTH_MIN = 2;
|
|
34 my $STAT_MIN_DEPTH_MAX = 10;
|
|
35 my $STAT_MIN_DEPTH_STEP = 2;
|
|
36 my $STAT_MAX_DEPTH_MIN = 100;
|
|
37 my $STAT_MAX_DEPTH_MAX = 200;
|
|
38 my $STAT_MAX_DEPTH_STEP = 100;
|
|
39 my $STAT_FREQ_MIN = 0.8;
|
|
40 my $STAT_FREQ_MAX = 1;
|
|
41 my $STAT_FREQ_STEP = 0.1;
|
|
42 my $STAT_DIST_MIN = 0;
|
|
43 my $STAT_DIST_MAX = 50;
|
|
44 my $STAT_DIST_STEP = 50;
|
|
45
|
|
46 GetOptions (
|
|
47 "input_file=s" => \$inputfile,
|
|
48 "log_file=s" => \$logfile,
|
|
49 "min_depth=i" => \$MIN_DEPTH,
|
|
50 "max_depth=i" => \$MAX_DEPTH,
|
|
51 "min_frequency=f" => \$MIN_VARIANTFREQUENCY,
|
|
52 "min_distance=i" => \$MIN_DISTANCE,
|
|
53 "min_forward_and_reverse=i" => \$MIN_FORWARDREVERSE,
|
|
54 "variant_only=s" => \$ONLY_UNFILTERED_VARIANT,
|
|
55 "v=i" => \$VERBOSE,
|
|
56 "do_stat=s" => \$DO_STAT,
|
|
57 "stat_min_depth_min=i" => \$STAT_MIN_DEPTH_MIN,
|
|
58 "stat_min_depth_max=i" => \$STAT_MIN_DEPTH_MAX,
|
|
59 "stat_min_depth_step=i" => \$STAT_MIN_DEPTH_STEP,
|
|
60 "stat_max_depth_min=i" => \$STAT_MAX_DEPTH_MIN,
|
|
61 "stat_max_depth_max=i" => \$STAT_MAX_DEPTH_MAX,
|
|
62 "stat_max_depth_step=i" => \$STAT_MAX_DEPTH_STEP,
|
|
63 "stat_freq_min=f" => \$STAT_FREQ_MIN,
|
|
64 "stat_freq_max=f" => \$STAT_FREQ_MAX,
|
|
65 "stat_freq_step=f" => \$STAT_FREQ_STEP,
|
|
66 "stat_dist_min=i" => \$STAT_DIST_MIN,
|
|
67 "stat_dist_max=i" => \$STAT_DIST_MAX,
|
|
68 "stat_dist_step=i" => \$STAT_DIST_STEP
|
|
69 ) or die("Error in command line arguments\n");
|
|
70
|
|
71 open(IF, $inputfile) or die("Can't open $inputfile\n");
|
|
72
|
|
73 my @tbl_line;
|
|
74 my %USR_PARAM;
|
|
75 $USR_PARAM{"min_depth"} = $MIN_DEPTH;
|
|
76 $USR_PARAM{"max_depth"} = $MAX_DEPTH;
|
|
77 $USR_PARAM{"min_freq"} = $MIN_VARIANTFREQUENCY;
|
|
78 $USR_PARAM{"min_dist"} = $MIN_DISTANCE;
|
|
79 $USR_PARAM{"min_fr"} = $MIN_FORWARDREVERSE;
|
|
80
|
|
81
|
|
82
|
|
83 #Extraction des variants
|
|
84 my $nb_line=0;
|
|
85 while (my $line=<IF>){
|
|
86 $nb_line++;
|
|
87 if (($nb_line % 1000000 == 0)&&($VERBOSE==1)){
|
|
88 print "$nb_line\n";
|
|
89 }
|
|
90 my $error_code=0;
|
|
91 if ($line=~/(.*?)\s+(\d+)\s+([ATGCN])\s+(\d+)\s+(.*?)\s+(.*?)$/){
|
|
92 my $current_chromosome = $1;
|
|
93 my $current_position = $2;
|
|
94 my $current_refbase = $3;
|
|
95 my $current_coverage = $4;
|
|
96 my $current_readbase_string = $5;
|
|
97 my $current_quality_string = $6;
|
|
98
|
|
99 #Suppression of mPileUp special character
|
|
100 $current_readbase_string =~ s/\$//g; #the read start at this position
|
|
101 $current_readbase_string =~ s/\^.//g; #the read end at this position followed by quality char
|
|
102
|
|
103 if ($current_readbase_string =~ /[ATGCNatgcn\d]/){
|
|
104 my %variant;
|
|
105 $variant{"line"} = $line;
|
|
106 $variant{"chr"} = $current_chromosome;
|
|
107 $variant{"pos"} = $current_position;
|
|
108 $variant{"refbase"} = $current_refbase;
|
|
109 $variant{"coverage"} = $current_coverage;
|
|
110 $variant{"readbase"} = $current_readbase_string;
|
|
111 $variant{"quality"} = $current_quality_string;
|
|
112 push(@tbl_line,\%variant);
|
|
113
|
|
114 if ($ONLY_UNFILTERED_VARIANT eq "ON"){
|
|
115 print $line;
|
|
116 }
|
|
117
|
|
118 }
|
|
119 else {
|
|
120 #Position with no variant
|
|
121 }
|
|
122
|
|
123 }
|
|
124 else {
|
|
125 #Error Parsing
|
|
126 print STDERR "$line #8";
|
|
127 }
|
|
128 }
|
|
129 close(IF);
|
|
130
|
|
131 if ($ONLY_UNFILTERED_VARIANT eq "ON"){
|
|
132 exit(0);
|
|
133 }
|
|
134
|
|
135 ####Checking the distance between variant and other filter
|
|
136
|
|
137
|
|
138 my @error;
|
|
139 for (my $i=0;$i<=$#tbl_line;$i++){
|
|
140 # print "ligne : $tbl_line[$i]\n";
|
|
141 my $before="";
|
|
142 my $after="";
|
|
143 my %line = %{$tbl_line[$i]};
|
|
144
|
|
145 if ($tbl_line[$i-1]){
|
|
146 $before = $tbl_line[$i-1];
|
|
147 }
|
|
148 if ($tbl_line[$i+1]){
|
|
149 $after = $tbl_line[$i+1];
|
|
150 }
|
|
151 my $error_code = check_error($tbl_line[$i],$before,$after,\%USR_PARAM);
|
|
152 if ($error_code == 0){
|
|
153 print $line{"line"};
|
|
154 }
|
|
155 else {
|
|
156 push(@error,$error_code,"\t",$line{"line"});
|
|
157 }
|
|
158 }
|
|
159
|
|
160 ### LOG
|
|
161 open(LF,">$logfile") or die ("Can't open $logfile\n");
|
|
162
|
|
163 if ($DO_STAT eq "YES"){
|
|
164 for (my $idx_min_depth=$STAT_MIN_DEPTH_MIN;$idx_min_depth<=$STAT_MIN_DEPTH_MAX;$idx_min_depth = $idx_min_depth + $STAT_MIN_DEPTH_STEP ){
|
|
165 for (my $idx_max_depth=$STAT_MAX_DEPTH_MIN;$idx_max_depth<=$STAT_MAX_DEPTH_MAX;$idx_max_depth = $idx_max_depth + $STAT_MAX_DEPTH_STEP ){
|
|
166 for (my $idx_freq = $STAT_FREQ_MIN;$idx_freq<=$STAT_FREQ_MAX;$idx_freq= $idx_freq+$STAT_FREQ_STEP){
|
|
167 for (my $idx_dist=$STAT_DIST_MIN;$idx_dist<=$STAT_DIST_MAX;$idx_dist = $idx_dist + $STAT_DIST_STEP){
|
|
168 for (my $idx_fr=0;$idx_fr<=1;$idx_fr++){
|
|
169 my %stat_param;
|
|
170 $stat_param{"min_depth"}=$idx_min_depth;
|
|
171 $stat_param{"max_depth"}=$idx_max_depth;
|
|
172 $stat_param{"min_freq"}=$idx_freq;
|
|
173 $stat_param{"min_fr"}=$idx_fr;
|
|
174 $stat_param{"min_dist"}=$idx_dist;
|
|
175
|
|
176 print LF "#SNP = ",&test_check(\@tbl_line,\%stat_param),"\tdepth (min/max) = ",$stat_param{"min_depth"}," / ",$stat_param{"max_depth"},"\tmin_dist=",$stat_param{"min_dist"},"\tmin_freq=",$stat_param{"min_freq"},"\tmin_forwardreverse = ",$stat_param{"min_fr"},"\n";
|
|
177 }
|
|
178 }
|
|
179 }
|
|
180 print "\n";
|
|
181 }
|
|
182 }
|
|
183 }
|
|
184
|
|
185
|
|
186 for (my $i=0;$i<=$#error;$i++){
|
|
187 print LF $error[$i];
|
|
188 }
|
|
189 close (LF);
|
|
190
|
|
191
|
|
192
|
|
193
|
|
194 sub test_check{
|
|
195 my $ref_tbl_line = shift;
|
|
196 my $ref_param = shift;
|
|
197 my @tbl_line = @$ref_tbl_line;
|
|
198 my %param = %$ref_param;
|
|
199 my $nb=0;
|
|
200
|
|
201 for (my $i=0;$i<=$#tbl_line;$i++){
|
|
202 my $before="";
|
|
203 my $after="";
|
|
204 my %line = %{$tbl_line[$i]};
|
|
205
|
|
206 if ($tbl_line[$i-1]){
|
|
207 $before = $tbl_line[$i-1];
|
|
208 }
|
|
209 if ($tbl_line[$i+1]){
|
|
210 $after = $tbl_line[$i+1];
|
|
211 }
|
|
212 my $error_code = check_error($tbl_line[$i],$before,$after,\%param);
|
|
213 if ($error_code == 0){
|
|
214 $nb++;
|
|
215 }
|
|
216 }
|
|
217
|
|
218 return $nb;
|
|
219 }
|
|
220
|
|
221 sub check_error{
|
|
222 my $refline = shift;
|
|
223 my %line = %$refline;
|
|
224 my $refbefore = shift;
|
|
225 my $refafter = shift;
|
|
226 my $refparam = shift;
|
|
227 my %param = %$refparam;
|
|
228
|
|
229
|
|
230 my $current_chromosome = $line{"chr"};
|
|
231 my $current_position = $line{"pos"};
|
|
232 my $current_refbase = $line{"refbase"};
|
|
233 my $current_coverage = $line{"coverage"};
|
|
234 my $current_readbase_string = $line{"readbase"};
|
|
235
|
|
236
|
|
237 my $min_depth = $param{"min_depth"};
|
|
238 my $max_depth = $param{"max_depth"};
|
|
239 my $min_variant_frequency = $param{"min_freq"};
|
|
240 my $min_forward_reverse = $param{"min_fr"};
|
|
241 my $min_dist = $param{"min_dist"};
|
|
242
|
|
243 #Verification of neightbourhood
|
|
244 if ($refbefore){
|
|
245 my %compareline = %$refbefore;
|
|
246 my $compare_chromosome = $compareline{"chr"};
|
|
247 my $compare_position = $compareline{"pos"};
|
|
248 my $compare_refbase = $compareline{"refbase"};
|
|
249 my $compare_coverage = $compareline{"coverage"};
|
|
250 my $compare_readbase_string = $compareline{"readbase"};
|
|
251
|
|
252 if (($current_chromosome eq $compare_chromosome )&&($compare_position + $min_dist >= $current_position)){
|
|
253 return 5;
|
|
254 }
|
|
255 }
|
|
256
|
|
257 if ($refafter){
|
|
258 my %compareline = %$refafter;
|
|
259 my $compare_chromosome = $compareline{"chr"};
|
|
260 my $compare_position = $compareline{"pos"};
|
|
261 my $compare_refbase = $compareline{"refbase"};
|
|
262 my $compare_coverage = $compareline{"coverage"};
|
|
263 my $compare_readbase_string = $compareline{"readbase"};
|
|
264
|
|
265 if (($current_chromosome eq $compare_chromosome )&&($current_position + $min_dist >= $compare_position)){
|
|
266 return 5;
|
|
267 }
|
|
268 }
|
|
269
|
|
270
|
|
271
|
|
272
|
|
273 #Extraction of insertions
|
|
274
|
|
275 ##################################################################
|
|
276 # my @IN = $current_readbase_string =~ m/\+[0-9]+[ACGTNacgtn]+/g;
|
|
277 # my @DEL = $current_readbase_string =~ m/\-[0-9]+[ACGTNacgtn]+/g;
|
|
278 # print "IN : @IN\n";
|
|
279 # print "DEL :@DEL\n";
|
|
280 #$current_readbase_string=~s/[\+\-][0-9]+[ACGTNacgtn]+//g;
|
|
281 ##################################################################
|
|
282 #!!! marche pas : exemple .+1Ct. correspond a . / +1C / t /. mais le match de l'expression vire +1Ct
|
|
283 ##################################################################
|
|
284
|
|
285 # => parcours de boucle
|
|
286 my @readbase = split(//,$current_readbase_string);
|
|
287 my $cleaned_readbase_string="";
|
|
288 my @IN;
|
|
289 my @DEL;
|
|
290 my $current_IN="";
|
|
291 my $current_DEL="";
|
|
292 my $current_size=0;
|
|
293
|
|
294 for (my $i=0;$i<=$#readbase;$i++){
|
|
295 if ($readbase[$i] eq "+"){
|
|
296 #Ouverture de IN
|
|
297 $current_IN="+";
|
|
298
|
|
299 #Recuperation de la taille
|
|
300 my $sub = substr $current_readbase_string,$i;
|
|
301 if ($sub=~/^\+(\d+)/){
|
|
302 $current_size = $1;
|
|
303 }
|
|
304 my $remaining_size = $current_size;
|
|
305 while (($remaining_size>0)&&($i<=$#readbase)){
|
|
306 $i++;
|
|
307 $current_IN.=$readbase[$i];
|
|
308 if ($readbase[$i]=~ /[ATGCNatgcn]/){
|
|
309 $remaining_size--;
|
|
310 }
|
|
311 }
|
|
312 push(@IN,$current_IN);
|
|
313 }
|
|
314 elsif ($readbase[$i] eq "-"){
|
|
315 #Ouverture de DEL
|
|
316 $current_DEL="-";
|
|
317
|
|
318 #Recuperation de la taille
|
|
319 my $sub = substr $current_readbase_string,$i;
|
|
320 if ($sub=~/^\-(\d+)/){
|
|
321 $current_size = $1;
|
|
322 }
|
|
323 my $remaining_size = $current_size;
|
|
324 while (($remaining_size>0)&&($i<=$#readbase)){
|
|
325 $i++;
|
|
326 $current_DEL.=$readbase[$i];
|
|
327 if ($readbase[$i]=~ /[ATGCNatgcn]/){
|
|
328 $remaining_size--;
|
|
329 }
|
|
330 }
|
|
331 push(@DEL,$current_DEL);
|
|
332
|
|
333 }
|
|
334 else {
|
|
335 #Ajout a la string
|
|
336 $cleaned_readbase_string .= $readbase[$i];
|
|
337 }
|
|
338 }
|
|
339
|
|
340
|
|
341 # print "IN : @IN\n";
|
|
342 # print "DEL :@DEL\n";
|
|
343 # print "$cleaned_readbase_string\n";
|
|
344
|
|
345 my @current_readbase_array = split(//,$cleaned_readbase_string);
|
|
346
|
|
347 #Filtering : error detection
|
|
348
|
|
349 if ($#current_readbase_array+1 != $current_coverage){
|
|
350 return 9;
|
|
351 #parsing error (couldn't parse the readbase string correctly)
|
|
352 }
|
|
353 elsif ($current_coverage<$min_depth){
|
|
354 return 2;
|
|
355 # 2 : insufficient depth
|
|
356 }
|
|
357 elsif ($current_coverage>$max_depth){
|
|
358 return 6;
|
|
359 # 6 : too much depth
|
|
360 }
|
|
361 else {
|
|
362 if ($#IN>=0){
|
|
363 if (($cleaned_readbase_string=~/[ACGTNacgtn]/)){
|
|
364 return 1;
|
|
365 # 1 : variant type overload (multiple variant type detected insertion/deletion/mutation)
|
|
366 }
|
|
367 else {
|
|
368 ########## TEST de coherence des insertions ################
|
|
369 # for (my $i=0;$i<=$#IN;$i++){
|
|
370 # if (uc($IN[0]) ne uc($IN[$i])){
|
|
371 # print uc($IN[0]),"\n";
|
|
372 # print uc($IN[$i]),"\n";
|
|
373 # return "1i";
|
|
374 # }
|
|
375 # }
|
|
376 ###########################################################
|
|
377
|
|
378 if($#IN+1 < $current_coverage*$min_variant_frequency ){
|
|
379 return 3;
|
|
380 # 3 : insufficient variant frequency
|
|
381 }
|
|
382 }
|
|
383 }
|
|
384 elsif ($#DEL>=0){
|
|
385 if (($cleaned_readbase_string=~/[ACGTNacgtn]/)){
|
|
386 return 1;
|
|
387 # 1 : variant type overload (multiple variant type detected insertion/deletion/mutation)
|
|
388 }
|
|
389 else {
|
|
390 ########## TEST de coherence des deletions ################
|
|
391 # for (my $i=0;$i<=$#DEL;$i++){
|
|
392 # if (uc($DEL[0]) ne uc($DEL[$i])){
|
|
393 # print uc($DEL[0]),"\n";
|
|
394 # print uc($DEL[$i]),"\n";
|
|
395 # return "1d";
|
|
396 # }
|
|
397 # }
|
|
398 ###########################################################
|
|
399
|
|
400 if($#DEL+1 < $current_coverage*$min_variant_frequency){
|
|
401 return 3;
|
|
402 # 3 : insufficient variant frequency
|
|
403 }
|
|
404 }
|
|
405 }
|
|
406 else {
|
|
407 my $nbA=0;
|
|
408 $nbA++ while ($current_readbase_string =~ m/A/g);
|
|
409 my $nbC=0;
|
|
410 $nbC++ while ($current_readbase_string =~ m/C/g);
|
|
411 my $nbT=0;
|
|
412 $nbT++ while ($current_readbase_string =~ m/T/g);
|
|
413 my $nbG=0;
|
|
414 $nbG++ while ($current_readbase_string =~ m/G/g);
|
|
415 my $nbN=0;
|
|
416 $nbN++ while ($current_readbase_string =~ m/N/g);
|
|
417 my $nba=0;
|
|
418 $nba++ while ($current_readbase_string =~ m/a/g);
|
|
419 my $nbc=0;
|
|
420 $nbc++ while ($current_readbase_string =~ m/c/g);
|
|
421 my $nbt=0;
|
|
422 $nbt++ while ($current_readbase_string =~ m/t/g);
|
|
423 my $nbg=0;
|
|
424 $nbg++ while ($current_readbase_string =~ m/g/g);
|
|
425 my $nbn=0;
|
|
426 $nbn++ while ($current_readbase_string =~ m/n/g);
|
|
427
|
|
428 if (($nbA+$nba>0)&&($nbT+$nbt+$nbG+$nbg+$nbC+$nbc+$nbN+$nbn>0)){
|
|
429 return "1m";
|
|
430 }
|
|
431 if (($nbT+$nbt>0)&&($nbA+$nba+$nbG+$nbg+$nbC+$nbc+$nbN+$nbn>0)){
|
|
432 return "1m";
|
|
433 }
|
|
434 if (($nbG+$nbg>0)&&($nbA+$nba+$nbT+$nbt+$nbC+$nbc+$nbN+$nbn>0)){
|
|
435 return "1m";
|
|
436 }
|
|
437 if (($nbC+$nbc>0)&&($nbA+$nba+$nbT+$nbt+$nbG+$nbg+$nbN+$nbn>0)){
|
|
438 return "1m";
|
|
439 }
|
|
440 if (($nbN+$nbn>0)&&($nbA+$nba+$nbT+$nbt+$nbG+$nbg+$nbC+$nbc>0)){
|
|
441 return "1m";
|
|
442 }
|
|
443
|
|
444 if ($nbA+$nba >= $current_coverage*$min_variant_frequency){
|
|
445 if (($nbA<$min_forward_reverse)||($nba<$min_forward_reverse)){
|
|
446 return 4;
|
|
447 # 4 : variant position not covered by forward and reverse reads
|
|
448 }
|
|
449 }
|
|
450 elsif ($nbT+$nbt >= $current_coverage*$min_variant_frequency){
|
|
451 if (($nbT<$min_forward_reverse)||($nbt<$min_forward_reverse)){
|
|
452 return 4;
|
|
453 # 4 : variant position not covered by forward and reverse reads
|
|
454 }
|
|
455 }
|
|
456 elsif ($nbG+$nbg >= $current_coverage*$min_variant_frequency){
|
|
457 if (($nbG<$min_forward_reverse)||($nbg<$min_forward_reverse)){
|
|
458 return 4;
|
|
459 # 4 : variant position not covered by forward and reverse reads
|
|
460 }
|
|
461 }
|
|
462 elsif ($nbC+$nbc >= $current_coverage*$min_variant_frequency){
|
|
463 if (($nbC<$min_forward_reverse)||($nbc<$min_forward_reverse)){
|
|
464 return 4;
|
|
465 # 4 : variant position not covered by forward and reverse reads
|
|
466 }
|
|
467 }
|
|
468 elsif ($nbN+$nbn >= $current_coverage*$min_variant_frequency){
|
|
469 if (($nbN<$min_forward_reverse)||($nbn<$min_forward_reverse)){
|
|
470 return 4;
|
|
471 # 4 : variant position not covered by forward and reverse reads
|
|
472 }
|
|
473 }
|
|
474 else {
|
|
475 return 3;
|
|
476 # 3 : insufficient variant frequency
|
|
477 }
|
|
478 }
|
|
479 }
|
|
480
|
|
481 return 0;
|
|
482 }
|