|
1
|
1 #!/usr/bin/perl -w
|
|
|
2
|
|
|
3 #Margaret Antonio 17.01.10 For GALAXY
|
|
|
4
|
|
|
5 use Data::Dumper;
|
|
|
6 use strict;
|
|
|
7 use Getopt::Long;
|
|
|
8 use Statistics::Distributions;
|
|
|
9 no warnings;
|
|
|
10
|
|
|
11 #ASSIGN INPUTS TO VARIABLES USING FLAGS
|
|
|
12 our ($input1, $input2, $help,$out,$round,$l1, $l2);
|
|
|
13
|
|
|
14 GetOptions(
|
|
|
15 'input1:s' => \$input1,
|
|
|
16 'input2:s' => \$input2,
|
|
|
17 'h' => \$help,
|
|
|
18 'o:s' =>\$out,
|
|
|
19 'r:i'=> \$round,
|
|
|
20 'l1:s'=> \$l1,
|
|
|
21 'l2:s'=> \$l2,
|
|
|
22 );
|
|
|
23
|
|
|
24 sub print_usage() {
|
|
|
25 print "\n####################################################################\n";
|
|
|
26 print "\n";
|
|
|
27 print "compWindows.pl: compare regions outputted by the slidingWindow tool\n\n";
|
|
|
28 print "DESCRIPTION: Takes two slidingWindows.csv files and compares them by\n";
|
|
|
29 print "calculating the difference in mean fitness, the pval for each gene.\n";
|
|
|
30 print "Example: compare control vs antibiotic, where both from same strain (genome).\n";
|
|
|
31 print "Note: For different strains/genomes, use compStrains for a gene comparison.pl\n";
|
|
|
32
|
|
|
33 print "\nUSAGE:\n";
|
|
|
34 print "perl compGenes.pl -d inputs/ -l cond1,cond2 -r 2 -o comp-cond1cond2_date.csv\n";
|
|
|
35
|
|
|
36 print "\nREQUIRED:\n";
|
|
|
37 print " -d\tDirectory containing two slidingWindow.csv files (output of\n";
|
|
|
38 print " \tslidingWindow tool)\n";
|
|
|
39 print " \tOR\n";
|
|
|
40 print " \tIn the command line (without a flag), input the name(s) of\n";
|
|
|
41 print " \ttwo files.\n";
|
|
|
42
|
|
|
43 print "\nOPTIONAL:\n";
|
|
|
44 print " -h\tPrints usage and exits program\n";
|
|
|
45 print " -o\tOutput file for comparison data. Default: label1label2.csv\n";
|
|
|
46 print " -r\tRound final output numbers to this number of decimals\n";
|
|
|
47 print " -s\tColumn number to sort by. Default: difference of means\n";
|
|
|
48 print " -l\tLabels for input files. Default: filenames\n";
|
|
|
49 print " \tTwo strings, comma separated (i.e. -l expt1,expt2).\n";
|
|
|
50 print " \tOrder should match file order.\n";
|
|
|
51
|
|
|
52 print " \n~~~~Always check that file paths are correctly specified~~~~\n";
|
|
|
53 print "\n##################################################################\n";
|
|
|
54 }
|
|
|
55 if ($help){
|
|
|
56 print_usage();
|
|
|
57 exit;
|
|
|
58 }
|
|
|
59
|
|
|
60
|
|
|
61
|
|
|
62 #THE @files ARRAY WILL CONTAIN INPUT FILE NAMES
|
|
|
63 my @files=($input1,$input2);
|
|
|
64
|
|
|
65 #GET LABELS: USE (-l) OR USE FILNEAMES AS LABELS FOR COLUMNS IN OUTPUT FILE
|
|
|
66
|
|
|
67 my @labels=($l1,$l2);
|
|
|
68
|
|
|
69 #INDICES 0:start 1:end 2:mutants 3:insertions 4:TA_sites 5:ratio 6:p-value
|
|
|
70 #7:average 8:variance 9:stdev 10:stderr 11:genes 12:fit-mean
|
|
|
71
|
|
|
72 #CHECK IF REQ. VARIABLES WERE DEFINED USING FLAGS. IF NOT THEN USE DEFAULT VALUES
|
|
|
73
|
|
|
74 if (!$out) {$out="comp.".$labels[0].$labels[1].".csv"}
|
|
|
75 if (!$round){$round='%.4f'}
|
|
|
76
|
|
|
77 #READ INPUT FILES AND PUT GENE INFO FROM EACH LINE INTO HASH %all WHERE KEY=GENE_ID AND
|
|
|
78 #VALUE=GENE INFO (MEAN FITNESS, # INSERTIONS, ETC.)
|
|
|
79
|
|
|
80 my %all;
|
|
|
81 my @header;
|
|
|
82 for (my $i=0; $i<2; $i++){
|
|
|
83 print "File #",$i+1,"\t";
|
|
|
84 my $file=$files[$i];
|
|
|
85 print $file,"\n";
|
|
|
86
|
|
|
87 open(DATA, '<', $file) or (print "Could not open '$file'\n" and print_usage() and exit);
|
|
|
88
|
|
|
89 #EXTRACT THE HEADER (COLUMN NAMES) OF THE FILE AND KEEP FOR OUTPUT HEADER
|
|
|
90 #APPEND FILE NAME OR GIVEN LABEL (-l) TO COLUMN NAME SO ITS DIFFERENT FROM OTHER INPUT FILE
|
|
|
91
|
|
|
92 my $head=<DATA>;
|
|
|
93 my @cols=split(',',$head);
|
|
|
94 for (my $j=0;$j<scalar @cols;$j++){
|
|
|
95 my $colname=$cols[$j];
|
|
|
96 chomp($colname);
|
|
|
97 $cols[$j]=$colname.'-'.$labels[$i];
|
|
|
98 print $cols[$j],"\n";
|
|
|
99 }
|
|
|
100 push (@header,@cols);
|
|
|
101
|
|
|
102 while (my $entry = <DATA>) {
|
|
|
103 chomp $entry;
|
|
|
104 my @line=split(",",$entry);
|
|
|
105 if (!$line[11]){
|
|
|
106 $line[11]="NA";
|
|
|
107 }
|
|
|
108 my ($start,$end,$mutants,$insertions,$TAsites,$ratio,$pval,$avg,$var,$sd,$se,$genes)=@line;
|
|
|
109 $avg=sprintf("%.3f",$avg);
|
|
|
110
|
|
|
111 #if (!$line[6]){
|
|
|
112 # $line[6]=0;
|
|
|
113 #}
|
|
|
114 #@line=@line[0,1,2,3,4,5,6];
|
|
|
115
|
|
|
116
|
|
|
117 #PUT GENE AND INFO INTO THE HASH FOR EXISTING KEY (1ST FILE) OR CREATE NEW KEY (2ND FILE)
|
|
|
118 my $key=$start;
|
|
|
119 if(!exists $all{$key}){
|
|
|
120 my @info;
|
|
|
121 push (@info,@line);
|
|
|
122 $all{$key}=\@info;
|
|
|
123 }
|
|
|
124 #Otherwise the window existed for the prior file, and now need to calcualte extra stats
|
|
|
125 else{
|
|
|
126 my @info=@{$all{$key}};
|
|
|
127 my ($start2,$end2,$mutants2,$insertions2,$TAsites2,$ratio2,$pval2,$avg2,$var2,$sd2,$se2,$genes2)=@info;
|
|
|
128 my $diff=sprintf("$round",($avg-$avg2));
|
|
|
129 my $df=$insertions+$insertions2-2;
|
|
|
130 my $tdist;
|
|
|
131 my $pval;
|
|
|
132
|
|
|
133 # CHECK TO MAKE SURE ALL VARIABLES IN TDIST,PVAL CALCULATIONS ARE NUMBERS AND NO
|
|
|
134 # ZEROS (0) IN THE DENOMINATOR
|
|
|
135
|
|
|
136 if ($se eq "NA" or $se2 eq "NA" or $sd eq "NA" or $sd2 eq "NA" or $insertions==0 or $insertions2==0 or $sd==0 or $sd2==0 or $df<=0){
|
|
|
137 ($tdist,$pval)=(" "," ");
|
|
|
138 }
|
|
|
139 else{
|
|
|
140 $tdist=sqrt((($diff)/(sqrt((($sd**2)/$insertions)+(($sd2**2)/$insertions2))))**2);
|
|
|
141 $pval=Statistics::Distributions::tprob($df,$tdist);
|
|
|
142 }
|
|
|
143 push (@info,@line,$diff,$df,$tdist,$pval);
|
|
|
144 $all{$key}=\@info;
|
|
|
145 }
|
|
|
146 }
|
|
|
147 close DATA;
|
|
|
148 }
|
|
|
149
|
|
|
150 #READ ALL HASH CONTENTS INTO 2D ARRAY FOR EASY SORTING AND PRINTING TO OUT FILE
|
|
|
151 my @unsorted;
|
|
|
152
|
|
|
153 foreach my $entry (keys %all) {
|
|
|
154 my @info=@{$all{$entry}};
|
|
|
155 my @temp;
|
|
|
156 push (@temp,@info);
|
|
|
157 push (@unsorted,\@temp);
|
|
|
158 }
|
|
|
159
|
|
|
160 #SORT GENES BY PVALUE OR FITNESS DEPENDING ON WHICH FLAG WAS USED IN COMMANDLINE
|
|
|
161
|
|
|
162 my $sortkey=26; #default: sort by difference of means
|
|
|
163
|
|
|
164 my @sorted = sort { $b->[$sortkey] <=> $a->[$sortkey] } @unsorted;
|
|
|
165
|
|
|
166 #ADD NEW FIELDS TO HEADER (COLUMN NAMES)
|
|
|
167 my $field="Mean".$labels[0].'.'.$labels[1];
|
|
|
168 push (@header,$field,"DOF","TDIST","PVALUE");
|
|
|
169
|
|
|
170 #PRINT TO OUT FILE
|
|
|
171 open OUT, '>',"$out";
|
|
|
172 print OUT (join(',',@header),"\n");
|
|
|
173 foreach (@sorted){
|
|
|
174 my @woo=@{$_};
|
|
|
175 print OUT join(',',@woo),"\n";
|
|
|
176 }
|
|
|
177 close OUT;
|
|
|
178
|
|
|
179
|
