From 74ceb35592278e2798d7a862b75238fd3e215847 Mon Sep 17 00:00:00 2001
From: Felix Krueger <fkrueger@altoslabs.com>
Date: Fri, 27 Dec 2024 20:41:17 +0100
Subject: [PATCH] Now works with FastA files and both options. Closes #723.

---
 CHANGELOG.md |  2 ++
 bismark      | 45 +++++++++++++++++++++++++++++++++++----------
 2 files changed, 37 insertions(+), 10 deletions(-)

diff --git a/CHANGELOG.md b/CHANGELOG.md
index 06392e3..9971072 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -17,6 +17,8 @@
 
 - now using 4 cores for merging multiple BAM files (more details [here](https://github.com/FelixKrueger/Bismark/issues/707)
 
+- fixed a corner case when reads were aligned in FastA mode with `--parallel` and in addition either `--ambiguous` and/or `--unmapped` (see [#723](https://github.com/FelixKrueger/Bismark/issues/723)
+
 ### deduplicate_bismark
 
 - Changed the path to Samtools to custom variable ([#609](https://github.com/FelixKrueger/Bismark/issues/609))
diff --git a/bismark b/bismark
index ed2b7fa..6d56841 100755
--- a/bismark
+++ b/bismark
@@ -596,7 +596,7 @@ foreach my $filename (@filenames){
 				}
 			}
 			if ($all_children_succeeded) {
-				print "All child process successfully finished.";
+				warn "All child processes successfully finished.";
 			}
 			else {
 				die "\nTerminating. Not all child processes successfully finished.";
@@ -743,23 +743,47 @@ foreach my $filename (@filenames){
 
 					if ($unmapped){
 						if ($gzip){
-							push @temp_unmapped_1,   "${output_dir}${add_prefix}${pathless_filename_1}.temp.${_}.gz_unmapped_reads_1.fq";
-							push @temp_unmapped_2,   "${output_dir}${add_prefix}${pathless_filename_2}.temp.${_}.gz_unmapped_reads_2.fq";
+							if ($sequence_file_format eq 'FASTA'){
+								push @temp_unmapped_1,   "${output_dir}${add_prefix}${pathless_filename_1}.temp.${_}.gz_unmapped_reads_1.fa";
+								push @temp_unmapped_2,   "${output_dir}${add_prefix}${pathless_filename_2}.temp.${_}.gz_unmapped_reads_2.fa";
+							}
+							else{
+								push @temp_unmapped_1,   "${output_dir}${add_prefix}${pathless_filename_1}.temp.${_}.gz_unmapped_reads_1.fq";
+								push @temp_unmapped_2,   "${output_dir}${add_prefix}${pathless_filename_2}.temp.${_}.gz_unmapped_reads_2.fq";
+							}
 						}
 						else{
-							push @temp_unmapped_1,   "${output_dir}${add_prefix}${pathless_filename_1}.temp.${_}_unmapped_reads_1.fq";
-							push @temp_unmapped_2,   "${output_dir}${add_prefix}${pathless_filename_2}.temp.${_}_unmapped_reads_2.fq";
+							if ($sequence_file_format eq 'FASTA'){
+								push @temp_unmapped_1,   "${output_dir}${add_prefix}${pathless_filename_1}.temp.${_}_unmapped_reads_1.fa";
+								push @temp_unmapped_2,   "${output_dir}${add_prefix}${pathless_filename_2}.temp.${_}_unmapped_reads_2.fa";
+							}
+							else{
+								push @temp_unmapped_1,   "${output_dir}${add_prefix}${pathless_filename_1}.temp.${_}_unmapped_reads_1.fq";
+								push @temp_unmapped_2,   "${output_dir}${add_prefix}${pathless_filename_2}.temp.${_}_unmapped_reads_2.fq";
+							}
 						}
 					}
 			  
 					if ($ambiguous){
 						if ($gzip){
-							push @temp_ambiguous_1,   "${output_dir}${add_prefix}${pathless_filename_1}.temp.${_}.gz_ambiguous_reads_1.fq";
-							push @temp_ambiguous_2,   "${output_dir}${add_prefix}${pathless_filename_2}.temp.${_}.gz_ambiguous_reads_2.fq";
+							if ($sequence_file_format eq 'FASTA'){
+								push @temp_ambiguous_1,   "${output_dir}${add_prefix}${pathless_filename_1}.temp.${_}.gz_ambiguous_reads_1.fa";
+								push @temp_ambiguous_2,   "${output_dir}${add_prefix}${pathless_filename_2}.temp.${_}.gz_ambiguous_reads_2.fa";
+							}
+							else{
+								push @temp_ambiguous_1,   "${output_dir}${add_prefix}${pathless_filename_1}.temp.${_}.gz_ambiguous_reads_1.fq";
+								push @temp_ambiguous_2,   "${output_dir}${add_prefix}${pathless_filename_2}.temp.${_}.gz_ambiguous_reads_2.fq";
+							}
 						}
 						else{
-							push @temp_ambiguous_1,   "${output_dir}${add_prefix}${pathless_filename_1}.temp.${_}_ambiguous_reads_1.fq";
-							push @temp_ambiguous_2,   "${output_dir}${add_prefix}${pathless_filename_2}.temp.${_}_ambiguous_reads_2.fq";
+							if ($sequence_file_format eq 'FASTA'){
+								push @temp_ambiguous_1,   "${output_dir}${add_prefix}${pathless_filename_1}.temp.${_}_ambiguous_reads_1.fa";
+								push @temp_ambiguous_2,   "${output_dir}${add_prefix}${pathless_filename_2}.temp.${_}_ambiguous_reads_2.fa";
+							}
+							else{
+								push @temp_ambiguous_1,   "${output_dir}${add_prefix}${pathless_filename_1}.temp.${_}_ambiguous_reads_1.fq";
+								push @temp_ambiguous_2,   "${output_dir}${add_prefix}${pathless_filename_2}.temp.${_}_ambiguous_reads_2.fq";
+							}
 						}
 					}
 				}
@@ -1262,7 +1286,8 @@ sub merge_individual_ambiguous_files{
   }
 
   foreach my $temp(@$temp_ambiguous){
-    $temp =~ s/.*\///; # removing path information
+	$temp =~ s/.*\///; # removing path information
+  	# warn "These are the temp ambiguous files: $temp\n"; sleep(5);
   }
 
   open (AMBIGUOUS,"| gzip -c - > $output_dir$ambiguous_file") or die "Failed to write to $ambiguous_file: $!\n";