use genome names and not genome hashes

when exporting gene aa sequences from genomes storages.
merenlab · Nov 20, 2023 · a62544f · a62544f
1 parent cf07d10
commit a62544f
Show file tree

Hide file tree

Showing 2 changed files with 6 additions and 3 deletions.
diff --git a/anvio/genomestorage.py b/anvio/genomestorage.py
@@ -436,7 +436,7 @@ def get_all_genome_names(self):
         return self.db.get_single_column_from_table(t.genome_info_table_name, 'genome_name')
 
 
-    def gen_combined_aa_sequences_FASTA(self, output_file_path, exclude_partial_gene_calls=False):
+    def gen_combined_aa_sequences_FASTA(self, output_file_path, exclude_partial_gene_calls=False, report_with_genome_name=False):
         self.run.info('Exclude partial gene calls', exclude_partial_gene_calls, nl_after=1)
 
         total_num_aa_sequences = 0
@@ -461,7 +461,10 @@ def gen_combined_aa_sequences_FASTA(self, output_file_path, exclude_partial_gene
 
                 aa_sequence = self.get_gene_sequence(genome_name, gene_caller_id)
 
-                fasta_output.write_id('%s_%d' % (genome_info_dict[genome_name]['genome_hash'], int(gene_caller_id)))
+                if report_with_genome_name:
+                    fasta_output.write_id('%s_%d' % (genome_name, int(gene_caller_id)))
+                else:
+                    fasta_output.write_id('%s_%d' % (genome_info_dict[genome_name]['genome_hash'], int(gene_caller_id)))
                 fasta_output.write_seq(aa_sequence, split=False)
 
                 total_num_aa_sequences += 1

diff --git a/bin/anvi-get-sequences-for-gene-calls b/bin/anvi-get-sequences-for-gene-calls
@@ -71,7 +71,7 @@ def export_from_genomes_storage(genomes_storage_db_path, output_file_path):
         raise ConfigError("GFF output is only relevant if you are working with a contigs database :/")
 
     g = genomestorage.GenomeStorage(genomes_storage_db_path)
-    g.gen_combined_aa_sequences_FASTA(output_file_path)
+    g.gen_combined_aa_sequences_FASTA(output_file_path, report_with_genome_name=True)
 
 
 def main(args):