make roary-like IGR presence absence csv file

tseemann · Nov 10, 2016 · 6ffd514 · 6ffd514
1 parent 362fdec
commit 6ffd514
Show file tree

Hide file tree

Showing 3 changed files with 26 additions and 15 deletions.
diff --git a/bin/.Rhistory b/bin/.Rhistory
@@ -0,0 +1,5 @@
+core_switched_region_divergences <- read.csv("/media/harry/extra/ST131/core_switched_region_divergences.csv", stringsAsFactors=FALSE)
+View(core_switched_region_divergences)
+library(cowplot)
+core_switched_region_divergences_subset <- core_switched_region_divergences[(core_switched_region_divergences$Nuc_identity < 0.8 & core_switched_region_divergences$Length_identity < 0.8), ]
+core_switched_region_divergences_subset <- core_switched_region_divergences[(core_switched_region_divergences$Nuc_identity < 0.8 | core_switched_region_divergences$Length_identity < 0.8), ]
diff --git a/bin/cluster_presence_absence.pl b/bin/cluster_presence_absence.pl
@@ -49,20 +49,21 @@
 
 @cluster_sorted_array=sort { $cluster_isolate_count_hash{$b} <=> $cluster_isolate_count_hash{$a} } keys %cluster_isolate_count_hash;
 
-print OUTPUT "Cluster,Isolates,Sequences";
+print OUTPUT "\"Gene\",\"Non-unique Gene name\",\"Annotation\",\"No. isolates\",\"No. sequences\",\"Avg sequences per isolate\",\"Genome Fragment\",\"Order within Fragment\",\"Accessory Fragment\",\"Accessory Order with Fragment\",\"QC\",\"Min group size nuc\",\"Max group size nuc\",\"Avg group size nuc\"";
 foreach $isolate(@isolate_array){
-	print OUTPUT ",$isolate";
+	print OUTPUT ",\"$isolate\"";
 }
 print OUTPUT "\n";
 
 foreach $cluster(@cluster_sorted_array){
+	$ave_seqs=($cluster_seq_count_hash{$cluster}/$cluster_isolate_count_hash{$cluster});
 
-	print OUTPUT "$cluster,$cluster_isolate_count_hash{$cluster},$cluster_seq_count_hash{$cluster}";
+	print OUTPUT "\"$cluster\",\"\",\"\",\"$cluster_isolate_count_hash{$cluster}\",\"$cluster_seq_count_hash{$cluster}\",\"$ave_seqs\",\"\",\"\",\"\",\"\",\"\",\"\",\"\",\"\"";
 	foreach $isolate(@isolate_array){
 		if($cluster_hash{$cluster}{$isolate}){
-			print OUTPUT ",$cluster_hash{$cluster}{$isolate}";
+			print OUTPUT ",\"$cluster_hash{$cluster}{$isolate}\"";
 		}else{
-			print OUTPUT ",";
+			print OUTPUT ",\"\"";
 		}
 	}
 	print OUTPUT "\n";

diff --git a/bin/roary_igry_combiner.pl b/bin/roary_igry_combiner.pl
@@ -1,8 +1,8 @@
 #!/usr/bin/perl -w
 
-use Text::CSV;
+#use Text::CSV;
 
-$csv = Text::CSV->new({ sep_char => ',', binary => 1 });
+#$csv = Text::CSV->new({ sep_char => ',', binary => 1 });
 
 $out_dir=$ARGV[0];
 $roary_dir=$ARGV[1];
@@ -13,12 +13,15 @@
 while(<INPUT_R>){
 	$line=$_;
 	$line=~s/\R//g;
+	$line=~s/^"//;
+	$line=~s/"$//;
+	@line_array=split(/","/, $line);
 
-	if($csv->parse($line)){
+	#if($csv->parse($line)){
 
-		@line_array=$csv->fields();
+	#	@line_array=$csv->fields();
 
-		if($line =~ /^"Gene",/){
+		if($line =~ /^Gene","/){
 			@header_array=@line_array;
 
 			$col_count=scalar(@line_array);
@@ -42,21 +45,23 @@
 				}
 			}
 		}
-	}
+	#}
 }
 
 open INPUT_I, "$out_dir/IGR_presence_absence.csv";
 while(<INPUT_I>){
 	$line=$_;
-	chomp $line;
-	@line_array=split(/,/, $line);
+	$line=~s/\R//g;
+	$line=~s/^"//;
+	$line=~s/"$//;
+	@line_array=split(/","/, $line);
 
-	if($line =~ /^Cluster,Isolates,/){
+	if($line =~ /^Gene","/){
 		@header_array=@line_array;
 
 		$col_count=scalar(@line_array);
 
-		$isolate_sta=3;
+		$isolate_sta=14;
 		$isolate_end=$col_count - 1;
 
 		#for($i=$isolate_sta; $i<=$isolate_end; $i++){