comparison compGenes.pl @ 11:21779cc1f7b6 draft

Uploaded
author antmarge
date Wed, 29 Mar 2017 19:41:57 -0400
parents
children
comparison
equal deleted inserted replaced
10:41673e6f67db 11:21779cc1f7b6
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, $outfile);
15
16 GetOptions(
17 'input1:s' => \$input1,
18 'input2:s' => \$input2,
19 'h' => \$h,
20 'o:s' =>\$outfile,
21 'r:i'=> \$round,
22 'l1:s'=>\$l1,
23 'l2:s'=>\$l2
24 );
25
26 sub print_usage() {
27 print "\n####################################################################\n";
28
29 print "compGenes: compare genes for an organism under different conditions\n\n";
30 print "DESCRIPTION: Takes two aggregate.pl outputs and compares them by\n";
31 print "calculating the difference in mean fitness for each gene.\n";
32 print "Example: compare organism in presence of control vs antibiotic.\n";
33 print "Note: For different strains/genomes, use compStrains.pl\n";
34
35 print "\nUSAGE:";
36 print "perl compGenes.pl -d inputs/ \n\n";
37
38 print "REQUIRED:\n";
39 print " -d\tDirectory containing all input files (files from\n";
40 print " \taggregate script)\n";
41 print " \tOR\n";
42 print " \tIn the command line (without a flag), input the name(s) of\n";
43 print " \ttwo files containing aggregate gene fitness values. \n\n";
44
45 print "OPTIONAL:\n";
46 print " -h\tPrints usage and exits program\n";
47 print " -o\tOutput file for comparison data. Default: label1label2.csv\n";
48 print " -r\tRound final output numbers to this number of decimals\n";
49 print " -l1\tLabels for input files. Default: filenames\n";
50 print " \tTwo strings, comma separated (i.e. -l expt1,expt2).\n";
51 print " \tOrder should match file order.\n";
52
53 print " \n~~~~Always check that file paths are correctly specified~~~~\n";
54 print "\n##################################################################\n";
55
56 }
57
58 # Check if help needed or if improper inputs
59
60 if ($h){
61 print_usage();
62 exit;
63 }
64
65 if (! $outfile){
66 $outfile="geneComp.csv"
67 }
68 $round = '%.4f';
69
70 my @files=($input1,$input2);
71
72 #GET LABELS: USE (-l) OR USE FILNEAMES AS LABELS FOR COLUMNS IN OUTPUT FILE
73
74 my @labels=($l1,$l2);
75
76 #CHECK IF REQ. VARIABLES WERE DEFINED USING FLAGS. IF NOT THEN USE DEFAULT VALUES
77
78 #$out="comp.".$labels[0].$labels[1].".csv";
79
80 #READ INPUT FILES AND PUT GENE INFO FROM EACH LINE INTO HASH %all WHERE KEY=GENE_ID AND
81 #VALUE=GENE INFO (MEAN FITNESS, # INSERTIONS, ETC.)
82
83 my %all;
84 my @header;
85
86 #OPEN TWO COMPARISON FILES, ONE PER LOOP
87 for (my $i=0; $i<2; $i++){
88 print "File #",$i+1,"\t";
89 my $file=$files[$i];
90 print $file,"\n";
91
92 open(DATA, '<', $file) or die "Could not open '$file'\n";
93
94 #EXTRACT THE HEADER (COLUMN NAMES) OF THE FILE AND KEEP FOR OUTPUT HEADER
95 #APPEND FILE NAME OR GIVEN LABEL (-l) TO COLUMN NAME SO ITS DIFFERENT FROM OTHER INPUT FILE
96
97 my $head=<DATA>;
98 my @temp = split("\n",$head);
99 $head = $temp[0];
100
101 my @cols=split(',',$head);
102 @cols = @cols[0,1,2,3,4,5];
103 for (my $j=0;$j<scalar @cols;$j++){
104 $cols[$j]=$cols[$j].'-'.$labels[$i];
105 }
106 push (@header,@cols);
107 while (my $entry = <DATA>) {
108 chomp $entry;
109 my @line=split(",",$entry);
110 if (!$line[5]){
111 $line[5]="NA";
112 }
113
114 @line=@line[0,1,2,3,4,5];
115 my $gene=$line[0];
116 chomp($gene);
117
118 #PUT GENE AND INFO INTO THE HASH FOR EXISTING KEY (1ST FILE) OR CREATE NEW KEY (2ND FILE)
119
120 if (!exists $all{$gene}){
121 my @info;
122 push (@info,@line);
123 $all{$gene}=\@info;
124 }
125 else{
126 my @info=@{$all{$gene}};
127 push (@info,@line);
128 my $diff=sprintf("$round",($info[1]-$info[7]));
129 my $total1=$info[2];
130 my $total2=$info[8];
131 my $sd1=$info[3];
132 my $se1=$info[4];
133 my $sd2=$info[9];
134 my $se2=$info[10];
135 my $df=$total1+$total2-2;
136 my $tdist;
137 my $pval;
138
139 # CHECK TO MAKE SURE ALL VARIABLES IN TDIST,PVAL CALCULATIONS ARE NUMBERS AND NO
140 # ZEROS (0) IN THE DENOMINATOR
141
142 if ($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 $df<=0){
143 ($tdist,$pval)=("NA","NA");
144 }
145 else{
146 $tdist=sqrt((($diff)/(sqrt((($sd1**2)/$total1)+(($sd2**2)/$total2))))**2);
147
148 $pval=Statistics::Distributions::tprob($df,$tdist);
149 }
150 push (@info,$diff,$df,$tdist,$pval);
151 $all{$gene}=\@info;
152 }
153 }
154 close DATA;
155 }
156
157 #READ ALL HASH CONTENTS INTO 2D ARRAY FOR EASY SORTING AND PRINTING TO OUT FILE
158 my @unsorted;
159
160 foreach my $entry (keys %all) {
161 my @info=@{$all{$entry}};
162 my @temp;
163 push (@temp,@info);
164 push (@unsorted,\@temp);
165 }
166
167 #SORT GENES BY PVALUE OR FITNESS DEPENDING ON WHICH FLAG WAS USED IN COMMANDLINE
168
169 $sortkey=15; #default: sort by difference of means
170
171 my @sorted = sort { $a->[$sortkey] <=> $b->[$sortkey] } @unsorted;
172
173 #ADD NEW FIELDS TO HEADER (COLUMN NAMES)
174 my $field="Mean".$labels[0].'.'.$labels[1];
175 push (@header,$field,"DOF","TDIST","PVALUE");
176
177 #PRINT TO OUT FILE
178 open OUT, '>',"$outfile";
179 print OUT (join(',',@header),"\n");
180 foreach (@sorted){
181 my @woo=@{$_};
182 print OUT join(',',@woo),"\n";
183 }
184 close OUT;
185
186