Mercurial > repos > antmarge > compgenes

comparison compGenes.pl @ 1:f3cbca89145f draft

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
Uploaded
author antmarge
date Tue, 28 Mar 2017 22:08:52 -0400
parents
children
comparison
equal deleted inserted replaced
0:ee05ebdd4879 1:f3cbca89145f
1 #!/usr/bin/perl -w
2
3 #Margaret Antonio 17.01.06 without essentiality. For original, unmodified aggregate.pl output (where unique insertions are not included).
4
5 use Getopt::Long;
6 use Statistics::Distributions;
7 use strict;
8 use autodie;
9 no warnings;
10
11 #no warnings;
12
13 #ASSIGN INPUTS TO VARIABLES USING FLAGS
14 our ($input1, $input2, $h, $out, $sortkey, $round, $l1, $l2, $outfile);
15
16 GetOptions(
17 'input1:s' => \$input1,
18 'input2:s' => \$input2,
19 'h' => \$h,
20 'o:s' =>\$outfile,
21 'r:i'=> \$round,
22
23 );
24
25 sub print_usage() {
26 print "\n####################################################################\n";
27
28 print "compGenes: compare genes for an organism under different conditions\n\n";
29 print "DESCRIPTION: Takes two aggregate.pl outputs and compares them by\n";
30 print "calculating the difference in mean fitness for each gene.\n";
31 print "Example: compare organism in presence of control vs antibiotic.\n";
32 print "Note: For different strains/genomes, use compStrains.pl\n";
33
34 print "\nUSAGE:";
35 print "perl compGenes.pl -d inputs/ \n\n";
36
37 print "REQUIRED:\n";
38 print " -d\tDirectory containing all input files (files from\n";
39 print " \taggregate script)\n";
40 print " \tOR\n";
41 print " \tIn the command line (without a flag), input the name(s) of\n";
42 print " \ttwo files containing aggregate gene fitness values. \n\n";
43
44 print "OPTIONAL:\n";
45 print " -h\tPrints usage and exits program\n";
46 print " -o\tOutput file for comparison data. Default: label1label2.csv\n";
47 print " -r\tRound final output numbers to this number of decimals\n";
48 print " -l1\tLabels for input files. Default: filenames\n";
49 print " \tTwo strings, comma separated (i.e. -l expt1,expt2).\n";
50 print " \tOrder should match file order.\n";
51
52 print " \n~~~~Always check that file paths are correctly specified~~~~\n";
53 print "\n##################################################################\n";
54
55 }
56
57 # Check if help needed or if improper inputs
58
59 if ($h){
60 print_usage();
61 exit;
62 }
63
64 if (! $outfile){
65 $outfile="geneComp.csv"
66 }
67 $round = '%.4f';
68
69 my @files=($input1,$input2);
70
71 #GET LABELS: USE (-l) OR USE FILNEAMES AS LABELS FOR COLUMNS IN OUTPUT FILE
72
73 my @labels=($l1,$l2);
74
75 #CHECK IF REQ. VARIABLES WERE DEFINED USING FLAGS. IF NOT THEN USE DEFAULT VALUES
76
77 #$out="comp.".$labels[0].$labels[1].".csv";
78
79 #READ INPUT FILES AND PUT GENE INFO FROM EACH LINE INTO HASH %all WHERE KEY=GENE_ID AND
80 #VALUE=GENE INFO (MEAN FITNESS, # INSERTIONS, ETC.)
81
82 my %all;
83 my @header;
84
85 #OPEN TWO COMPARISON FILES, ONE PER LOOP
86 for (my $i=0; $i<2; $i++){
87 print "File #",$i+1,"\t";
88 my $file=$files[$i];
89 print $file,"\n";
90
91 open(DATA, '<', $file) or die "Could not open '$file'\n";
92
93 #EXTRACT THE HEADER (COLUMN NAMES) OF THE FILE AND KEEP FOR OUTPUT HEADER
94 #APPEND FILE NAME OR GIVEN LABEL (-l) TO COLUMN NAME SO ITS DIFFERENT FROM OTHER INPUT FILE
95
96 my $head=<DATA>;
97 my @temp = split("\n",$head);
98 $head = $temp[0];
99
100 my @cols=split(',',$head);
101 @cols = @cols[0,1,2,3,4,5];
102 for (my $j=0;$j<scalar @cols;$j++){
103 $cols[$j]=$cols[$j].'-'.$labels[$i];
104 }
105 push (@header,@cols);
106 while (my $entry = <DATA>) {
107 chomp $entry;
108 my @line=split(",",$entry);
109 if (!$line[5]){
110 $line[5]="NA";
111 }
112 if (!$line[6]){
113 $line[6]=0;
114 }
115 @line=@line[0,1,2,3,4,5];
116 my $gene=$line[0];
117 chomp($gene);
118
119 #PUT GENE AND INFO INTO THE HASH FOR EXISTING KEY (1ST FILE) OR CREATE NEW KEY (2ND FILE)
120
121 if (!exists $all{$gene}){
122 my @info;
123 push (@info,@line);
124 $all{$gene}=\@info;
125 }
126 else{
127 my @info=@{$all{$gene}};
128 push (@info,@line);
129 my $diff=sprintf("$round",($info[1]-$info[7]));
130 my $total1=$info[2];
131 my $total2=$info[8];
132 my $sd1=$info[3];
133 my $se1=$info[4];
134 my $sd2=$info[9];
135 my $se2=$info[10];
136 my $df=$total1+$total2-2;
137 my $tdist;
138 my $pval;
139
140 # CHECK TO MAKE SURE ALL VARIABLES IN TDIST,PVAL CALCULATIONS ARE NUMBERS AND NO
141 # ZEROS (0) IN THE DENOMINATOR
142
143 if ($se1 eq "X" or $se2 eq "X" or $sd1 eq "X" or $sd2 eq "X" or $total1==0 or $total2==0 or $sd1==0 or $sd2==0 or $df<=0){
144 ($tdist,$pval)=("NA","NA");
145 }
146 else{
147 $tdist=sqrt((($diff)/(sqrt((($sd1**2)/$total1)+(($sd2**2)/$total2))))**2);
148
149 $pval=Statistics::Distributions::tprob($df,$tdist);
150 }
151 push (@info,$diff,$df,$tdist,$pval);
152 $all{$gene}=\@info;
153 }
154 }
155 close DATA;
156 }
157
158 #READ ALL HASH CONTENTS INTO 2D ARRAY FOR EASY SORTING AND PRINTING TO OUT FILE
159 my @unsorted;
160
161 foreach my $entry (keys %all) {
162 my @info=@{$all{$entry}};
163 my @temp;
164 push (@temp,@info);
165 push (@unsorted,\@temp);
166 }
167
168 #SORT GENES BY PVALUE OR FITNESS DEPENDING ON WHICH FLAG WAS USED IN COMMANDLINE
169
170 $sortkey=15; #default: sort by difference of means
171
172 my @sorted = sort { $a->[$sortkey] <=> $b->[$sortkey] } @unsorted;
173
174 #ADD NEW FIELDS TO HEADER (COLUMN NAMES)
175 my $field="Mean".$labels[0].'.'.$labels[1];
176 push (@header,$field,"DOF","TDIST","PVALUE");
177
178 #PRINT TO OUT FILE
179 open OUT, '>',"$outfile";
180 print OUT (join(',',@header),"\n");
181 foreach (@sorted){
182 my @woo=@{$_};
183 print OUT join(',',@woo),"\n";
184 }
185 close OUT;
186
187