ORForise 1.6.2__py3-none-any.whl

ORForise/Aggregate_Compare.py

@@ -0,0 +1,378 @@
+from importlib import import_module
+import argparse
+import csv, os, gzip, sys
+
+try:
+    from .Comparator import tool_comparison
+    from .utils import *
+except (ImportError, ModuleNotFoundError):
+    from Comparator import tool_comparison
+    from utils import *
+############################################
+
+def comparator(options):
+    try:
+        try:  # Detect whether fasta/gff files are .gz or text and read accordingly
+            fasta_in = gzip.open(options.genome_dna, 'rt')
+            dna_regions = fasta_load(fasta_in)
+        except:
+            fasta_in = open(options.genome_dna, 'r', encoding='unicode_escape')
+            dna_regions = fasta_load(fasta_in)
+        try:
+            gff_in = gzip.open(options.reference_annotation, 'rt')
+            dna_regions = gff_load(options, gff_in, dna_regions)
+        except:
+            gff_in = open(options.reference_annotation, 'r', encoding='unicode_escape')
+            dna_regions = gff_load(options, gff_in, dna_regions)
+    except AttributeError:
+        sys.exit("Attribute Error:\nStORF'ed GFF probably already exists - Must be deleted before running (-overwrite)")
+    except FileNotFoundError:
+        split_path = options.gff.split(os.sep)
+        sys.exit("Directory '" + split_path[-2] + "' missing fna/gff files")
+        ###############################################
+    total_ref_genes = sum(
+        len(v[2]) if isinstance(v[2], (list, tuple, set, dict, str)) else 1 for v in dna_regions.values())
+    #############################################
+    # Collect predictions from tools
+    aggregate_Predictions = collections.OrderedDict()
+    aggregate_Tools = options.tools.split(',')
+    for i, (tool) in enumerate(aggregate_Tools):
+        tool_prediction = options.tool_predictions.split(',')[i]
+        print(tool)
+        try:
+            tool_ = import_module('Tools.' + tool + '.' + tool, package='my_current_pkg')
+        except ModuleNotFoundError:
+            try:
+                tool_ = import_module('ORForise.Tools.' + tool + '.' + tool, package='my_current_pkg')
+            except ModuleNotFoundError:
+                sys.exit("Tool not available")
+        tool_ = getattr(tool_, tool)
+        ##
+        orfs = tool_(tool_prediction, dna_regions)
+        for current_contig in orfs:
+            if current_contig not in aggregate_Predictions:
+                aggregate_Predictions[current_contig] = {}
+            current_orfs = orfs[current_contig]
+            for key, value in current_orfs.items():
+                if key in aggregate_Predictions[current_contig]:
+                    aggregate_Predictions[current_contig][key][-1] += '|' + tool
+                else:
+                    aggregate_Predictions[current_contig][key] = value
+
+    aggregate_ORFs = {k: sortORFs(v) for k, v in aggregate_Predictions.items()}
+    results = tool_comparison(aggregate_ORFs, dna_regions, options.verbose)
+    ############## Printing to std-out and optional csv file
+    # Ensure the output directory exists
+    os.makedirs(options.outdir, exist_ok=True)
+    # Use outname as a directory, basename for files is output-outname
+    base_out = os.path.join(options.outdir, f"{os.path.basename(options.outname)}")
+
+    # Prepare to collect summary stats for all contigs
+    contig_summaries = []
+    ############################################# To get default output filename from input file details
+    if options.outdir:
+        # Ensure the output directory exists
+        os.makedirs(options.outdir, exist_ok=True)
+        # Use outname as a directory, basename for files is output-outname
+        base_out = os.path.join(options.outdir, f"{os.path.basename(options.outname)}")
+        with open(f"{base_out}_summary.txt", 'w', encoding='utf-8') as out_file:
+            out_file.write('Genome Used: ' + str(options.genome_dna.split('/')[-1]) + '\n')
+            if options.reference_tool:
+                out_file.write('Reference Tool Used: ' + str(options.reference_tool) + '\n')
+            else:
+                out_file.write('Reference Used: ' + str(options.reference_annotation.split('/')[-1]) + '\n')
+            out_file.write('Tool Compared: ' + str(options.tools) + '\n')
+            out_file.write('Total Number of Reference Genes: ' + str(total_ref_genes) + '\n')
+            out_file.write('Number of Contigs: ' + str(len(dna_regions)) + '\n')
+            out_file.write(
+                'Contig\tGenes\tORFs\tPerfect_Matches\tPartial_Matches\tMissed_Genes\tUnmatched_ORFs\tMulti_Matched_ORFs\n')
+
+    for dna_region, result in results.items():
+        num_current_genes = len(dna_regions[dna_region][2])
+        num_orfs = result['pred_metrics']['Number_of_ORFs']
+        num_perfect = result['pred_metrics']['Number_of_Perfect_Matches']
+        num_partial = len(result['pred_metrics']['partial_Hits'])
+        num_missed = len(result['rep_metrics']['genes_Undetected'])
+        num_unmatched = len(result['pred_metrics']['unmatched_ORFs'])
+        num_multi = len(result['pred_metrics']['multi_Matched_ORFs'])
+
+        ####
+        # Tool-specific stats
+        tool_stats = {}
+        for tool in options.tools.split(','):
+            tool_stats[tool] = {
+                'perfect': 0,
+                'partial': 0,
+                'unmatched': 0,
+                'multi': 0
+            }
+        # Count perfect matches per tool
+        for key in result['pred_metrics'].get('perfect_Matches', {}):
+            for tool in options.tools.split(','):
+                if tool in key:
+                    tool_stats[tool]['perfect'] += 1
+        # Count partial matches per tool
+        for key in result['pred_metrics'].get('partial_Hits', {}):
+            for tool in options.tools.split(','):
+                if tool in key:
+                    tool_stats[tool]['partial'] += 1
+        # Count unmatched ORFs per tool
+        for key in result['pred_metrics'].get('unmatched_ORFs', {}):
+            for tool in options.tools.split(','):
+                if tool in key:
+                    tool_stats[tool]['unmatched'] += 1
+        # Count multi-matched ORFs per tool
+        for key in result['pred_metrics'].get('multi_Matched_ORFs', {}):
+            for tool in options.tools.split(','):
+                if tool in key:
+                    tool_stats[tool]['multi'] += 1
+        ####
+
+        # Collect summary for this contig
+        if options.outdir:
+            contig_summaries.append([
+                dna_region, num_current_genes, num_orfs, num_perfect, num_partial, num_missed, num_unmatched, num_multi
+            ])
+        ###
+        num_current_genes = len(dna_regions[dna_region][2])
+        print("These are the results for: " + dna_region + '\n')
+        ############################################# To get default output filename from input file details
+        genome_name = options.reference_annotation.split('/')[-1].split('.')[0]
+        rep_metric_description, rep_metrics = get_rep_metrics(result)
+        all_metric_description, all_metrics = get_all_metrics(result)
+
+        print('Current Contig: ' + str(dna_region))
+        print('Number of Genes: ' + str(num_current_genes))
+        print('Number of ORFs: ' + str(result['pred_metrics']['Number_of_ORFs']))
+        print('Perfect Matches: ' + str(result['pred_metrics']['Number_of_Perfect_Matches']) + ' [' + str(num_current_genes)+ '] - '+ format(100 * result['pred_metrics']['Number_of_Perfect_Matches']/num_current_genes,'.2f')+'%')
+        print('Partial Matches: ' + str(len(result['pred_metrics']['partial_Hits'])) + ' [' + str(num_current_genes)+ '] - '+ format(100 * len(result['pred_metrics']['partial_Hits'])/num_current_genes,'.2f')+'%')
+        print('Missed Genes: ' + str(len(result['rep_metrics']['genes_Undetected'])) + ' [' + str(num_current_genes)+ '] - '+ format(100 * len(result['rep_metrics']['genes_Undetected'])/num_current_genes,'.2f')+'%')
+        print('Unmatched ORFs: ' + str(len(result['pred_metrics']['unmatched_ORFs'])) + ' [' + str(num_current_genes)+ '] - '+ format(100 * len(result['pred_metrics']['unmatched_ORFs'])/num_current_genes,'.2f')+'%')
+        print('Multi-matched ORFs: ' + str(len(result['pred_metrics']['multi_Matched_ORFs'])) + ' [' + str(num_current_genes)+ '] - '+ format(100 * len(result['pred_metrics']['multi_Matched_ORFs'])/num_current_genes,'.2f')+'%')
+        print('Tool breakdown:')
+        for tool, stats in tool_stats.items():
+            print(
+                f"  {tool}: Perfect={stats['perfect']}, Partial={stats['partial']}, Unmatched={stats['unmatched']}, Multi-matched={stats['multi']}")
+
+        if options.outdir:
+            # Prepare output directory and file names for each contig
+            contig_save = dna_region.replace('/', '_').replace('\\', '_')
+            contig_dir = os.path.join(options.outdir, contig_save)
+            os.makedirs(contig_dir, exist_ok=True)
+            summary_file = os.path.join(contig_dir, "summary.txt")
+            csv_file = os.path.join(contig_dir, "metrics.csv")
+            perfect_fasta = os.path.join(contig_dir, "perfect_matches.fasta")
+            partial_fasta = os.path.join(contig_dir, "partial_matches.fasta")
+            missed_fasta = os.path.join(contig_dir, "missed_genes.fasta")
+            unmatched_fasta = os.path.join(contig_dir, "unmatched_orfs.fasta")
+            multi_fasta = os.path.join(contig_dir, "multi_matched_orfs.fasta")
+
+            # Write summary to text file
+            with open(summary_file, 'w', encoding='utf-8') as sf:
+                sf.write('Current Contig: ' + str(dna_region) + '\n')
+                sf.write('Number of Genes: ' + str(num_current_genes) + '\n')
+                sf.write('Number of ORFs: ' + str(result['pred_metrics']['Number_of_ORFs']) + '\n')
+                sf.write('Perfect Matches: ' + str(result['pred_metrics']['Number_of_Perfect_Matches']) + ' [' + str(
+                    num_current_genes) + '] - ' + format(
+                    100 * result['pred_metrics']['Number_of_Perfect_Matches'] / num_current_genes, '.2f') + '%\n')
+                sf.write('Partial Matches: ' + str(len(result['pred_metrics']['partial_Hits'])) + ' [' + str(
+                    num_current_genes) + '] - ' + format(
+                    100 * len(result['pred_metrics']['partial_Hits']) / num_current_genes, '.2f') + '%\n')
+                sf.write('Missed Genes: ' + str(len(result['rep_metrics']['genes_Undetected'])) + ' [' + str(
+                    num_current_genes) + '] - ' + format(
+                    100 * len(result['rep_metrics']['genes_Undetected']) / num_current_genes, '.2f') + '%\n')
+                sf.write('Unmatched ORFs: ' + str(len(result['pred_metrics']['unmatched_ORFs'])) + ' [' + str(
+                    num_current_genes) + '] - ' + format(
+                    100 * len(result['pred_metrics']['unmatched_ORFs']) / num_current_genes, '.2f') + '%\n')
+                sf.write('Multi-matched ORFs: ' + str(len(result['pred_metrics']['multi_Matched_ORFs'])) + ' [' + str(
+                    num_current_genes) + '] - ' + format(
+                    100 * len(result['pred_metrics']['multi_Matched_ORFs']) / num_current_genes, '.2f') + '%\n')
+                sf.write('Tool breakdown:\n')
+                for tool, stats in tool_stats.items():
+                    sf.write(
+                        f"  {tool}: Perfect={stats['perfect']}, Partial={stats['partial']}, Unmatched={stats['unmatched']}, Multi-matched={stats['multi']}\n")
+
+            # Write metrics to CSV
+            with open(csv_file, 'w', newline='\n', encoding='utf-8') as out_file:
+                tool_out = csv.writer(out_file, quoting=csv.QUOTE_NONE, escapechar=" ")
+                tool_out.writerow(['Representative_Metrics:'])
+                tool_out.writerow(rep_metric_description.split(','))
+                tool_out.writerow([*rep_metrics])
+                tool_out.writerow(['Prediction_Metrics:'])
+                tool_out.writerow(all_metric_description.split(','))
+                tool_out.writerow([*all_metrics])
+                tool_out.writerow(['Reference_CDS_Gene_Coverage_of_Genome'])
+                tool_out.writerow([''.join(map(str, result['rep_metrics']['gene_Coverage_Genome']))])
+                tool_out.writerow(['Predicted_CDS_Coverage_of_Genome'])
+                tool_out.writerow([''.join(map(str, result['pred_metrics']['orf_Coverage_Genome']))])
+                tool_out.writerow(['Matched_Predicted_CDS_Coverage_of_Genome'])
+                tool_out.writerow([''.join(map(str, result['pred_metrics']['matched_ORF_Coverage_Genome']))])
+                # tool_out.writerow(['Start_Position_Difference:'])
+                # tool_out.writerow(result.get('start_Difference', []))
+                # tool_out.writerow(['Stop_Position_Difference:'])
+                # tool_out.writerow(result.get('stop_Difference', []))
+                # tool_out.writerow(['Alternative_Starts_Predicted:'])
+                # tool_out.writerow(result.get('other_Starts', []))
+                # tool_out.writerow(['Alternative_Stops_Predicted:'])
+                # tool_out.writerow(result.get('other_Stops', []))
+                # tool_out.writerow(['Undetected_Gene_Metrics:'])
+                # tool_out.writerow([
+                #     'ATG_Start,GTG_Start,TTG_Start,ATT_Start,CTG_Start,Alternative_Start_Codon,TGA_Stop,TAA_Stop,TAG_Stop,Alternative_Stop_Codon,Median_Length,ORFs_on_Positive_Strand,ORFs_on_Negative_Strand'
+                # ])
+                # tool_out.writerow(result.get('undetected_Gene_Metrics', []))
+                # tool_out.writerow(['\nPredicted_CDSs_Without_Corresponding_Gene_In_Reference_Metrics:'])
+                # tool_out.writerow([
+                #     'ATG_Start,GTG_Start,TTG_Start,ATT_Start,CTG_Start,Alternative_Start_Codon,TGA_Stop,TAA_Stop,TAG_Stop,Alternative_Stop_Codon,Median_Length,ORFs_on_Positive_Strand,ORFs_on_Negative_Strand'
+                # ])
+                # tool_out.writerow(result.get('unmatched_ORF_Metrics', []))
+
+            # Write perfect matches to FASTA
+            with open(perfect_fasta, 'w', encoding='utf-8') as f:
+                for key, value in result['pred_metrics'].get('perfect_Matches', {}).items():
+                    key_parts = key.split(',')
+                    id = f">{genome_name}_{key_parts[0]}_{key_parts[1]}_{key_parts[2]}_{key_parts[5]}"
+                    f.write(f"{id}\n{value}\n")
+
+            # Write partial matches to FASTA
+            with open(partial_fasta, 'w', encoding='utf-    8') as f:
+                for key, value in result['pred_metrics'].get('partial_Hits', {}).items():
+                    key_parts = key.split(';')
+                    gene_Seq = value[0]
+                    orf_Seq = value[1]
+                    f.write(f">{key_parts[0]}_gene\n{gene_Seq}\n>{key_parts[1]}_orf\n{orf_Seq}\n")
+
+            # Write missed genes to FASTA
+            with open(missed_fasta, 'w', encoding='utf-8') as f:
+                for key, value in result['rep_metrics'].get('genes_Undetected', {}).items():
+                    key_parts = key.split(',')
+                    id = f">{genome_name}_{key_parts[0]}_{key_parts[1]}_{key_parts[2]}"
+                    f.write(f"{id}\n{value}\n")
+
+            # Write unmatched ORFs to FASTA
+            with open(unmatched_fasta, 'w', encoding='utf-8') as f:
+                for key, value in result['pred_metrics'].get('unmatched_ORFs', {}).items():
+                    key_parts = key.split(',')
+                    id = f">{options.tools}_{key_parts[0]}_{key_parts[1]}_{key_parts[2]}"
+                    f.write(f"{id}\n{value}\n")
+
+            # Write multi-matched ORFs to FASTA
+            with open(multi_fasta, 'w', encoding='utf-8') as f:
+                for key, value in result['pred_metrics'].get('multi_Matched_ORFs', {}).items():
+                    key_parts = key.split(',')
+                    multi = f">Predicted_CDS:{key_parts[0]}-{key_parts[1]}_Genes:{'|'.join(value)}"
+                    f.write(f"{multi}\n")
+
+    # After all contigs, append the summary table to the main summary file
+    if options.outdir and contig_summaries:
+        with open(f"{base_out}_summary.txt", 'a', encoding='utf-8') as out_file:
+            for row in contig_summaries:
+                out_file.write('\t'.join(map(str, row)) + '\n')
+            # Optionally, add overall totals
+            total_genes = sum(row[1] for row in contig_summaries)
+            total_orfs = sum(row[2] for row in contig_summaries)
+            total_perfect = sum(row[3] for row in contig_summaries)
+            total_partial = sum(row[4] for row in contig_summaries)
+            total_missed = sum(row[5] for row in contig_summaries)
+            total_unmatched = sum(row[6] for row in contig_summaries)
+            total_multi = sum(row[7] for row in contig_summaries)
+            out_file.write('\nOverall Summary:\n')
+            out_file.write(f'Number of Genes: {total_genes}\n')
+            out_file.write(f'Number of ORFs: {total_orfs}\n')
+            out_file.write(
+                f'Perfect Matches: {total_perfect} [{total_genes}] - {format(100 * total_perfect / total_genes, ".2f")}%\n')
+            out_file.write(
+                f'Partial Matches: {total_partial} [{total_genes}] - {format(100 * total_partial / total_genes, ".2f")}%\n')
+            out_file.write(
+                f'Missed Genes: {total_missed} [{total_genes}] - {format(100 * total_missed / total_genes, ".2f")}%\n')
+            out_file.write(
+                f'Unmatched ORFs: {total_unmatched} [{total_genes}] - {format(100 * total_unmatched / total_genes, ".2f")}%\n')
+            out_file.write(
+                f'Multi-matched ORFs: {total_multi} [{total_genes}] - {format(100 * total_multi / total_genes, ".2f")}%\n')
+
+            # Calculate combined tool stats - could be optimised further
+            combined_tool_stats = {tool: {'perfect': 0, 'partial': 0, 'unmatched': 0, 'multi': 0} for tool in
+                                   options.tools.split(',')}
+            for dna_region, result in results.items():
+                for tool in options.tools.split(','):
+                    # perfect
+                    for key in result['pred_metrics'].get('perfect_Matches', {}):
+                        if tool in key:
+                            combined_tool_stats[tool]['perfect'] += 1
+                    # partial
+                    for key in result['pred_metrics'].get('partial_Hits', {}):
+                        if tool in key:
+                            combined_tool_stats[tool]['partial'] += 1
+                    # unmatched
+                    for key in result['pred_metrics'].get('unmatched_ORFs', {}):
+                        if tool in key:
+                            combined_tool_stats[tool]['unmatched'] += 1
+                    # multi
+                    for key in result['pred_metrics'].get('multi_Matched_ORFs', {}):
+                        if tool in key:
+                            combined_tool_stats[tool]['multi'] += 1
+            for tool, stats in combined_tool_stats.items():
+                out_file.write('\n'+
+                    f"  {tool}: Perfect={stats['perfect']}, Partial={stats['partial']}, Unmatched={stats['unmatched']}, Multi-matched={stats['multi']}\n"
+                )
+
+            # Print combined metrics to stdout
+            print("\nCombined metrics for all contigs:")
+            print(f'Number of Genes: {total_genes}')
+            print(f'Number of ORFs: {total_orfs}')
+            print(
+                f'Perfect Matches: {total_perfect} [{total_genes}] - {format(100 * total_perfect / total_genes, ".2f")}%')
+            print(
+                f'Partial Matches: {total_partial} [{total_genes}] - {format(100 * total_partial / total_genes, ".2f")}%')
+            print(f'Missed Genes: {total_missed} [{total_genes}] - {format(100 * total_missed / total_genes, ".2f")}%')
+            print(
+                f'Unmatched ORFs: {total_unmatched} [{total_genes}] - {format(100 * total_unmatched / total_genes, ".2f")}%')
+            print(
+                f'Multi-matched ORFs: {total_multi} [{total_genes}] - {format(100 * total_multi / total_genes, ".2f")}%')
+
+            print('Tool breakdown (combined):')
+            for tool, stats in combined_tool_stats.items():
+                print('\n'+
+                    f"  {tool}: Perfect={stats['perfect']}, Partial={stats['partial']}, Unmatched={stats['unmatched']}, Multi-matched={stats['multi']}"
+                )
+
+
+def main():
+    print(WELCOME)
+
+    parser = argparse.ArgumentParser(description='ORForise ' + ORForise_Version + ': Aggregate-Compare Run Parameters.')
+    parser._action_groups.pop()
+
+    required = parser.add_argument_group('Required Arguments')
+
+    required.add_argument('-dna', dest='genome_dna', required=True, help='Genome DNA file (.fa) which both annotations '
+                                                                    'are based on')
+    required.add_argument('-t', dest='tools', required=True, help='Which tools to analyse?')
+    required.add_argument('-tp', dest='tool_predictions', required=True, help='Tool genome prediction file (.gff) - Provide'
+                                                                         'file locations for each tool comma separated')
+    required.add_argument('-ref', dest='reference_annotation', required=True,
+                        help='Which reference annotation file to use as reference?')
+
+    optional = parser.add_argument_group('Optional Arguments')
+    optional.add_argument('-gene_ident', action='store', dest='gene_ident', default='CDS',
+                          help='What features to consider as genes? - Default: CDS - '
+                               'Provide comma separated list of features to consider as genes (e.g. CDS,exon)')
+    optional.add_argument('-rt', dest='reference_tool', required=False,
+                        help='What type of Annotation to compare to? -- Leave blank for Ensembl reference'
+                             '- Provide tool name to compare output from two tools')
+
+    output = parser.add_argument_group('Output')
+    output.add_argument('-o', dest='outdir', required=False,
+                        help='Define directory where detailed output should be places - If not provided, summary will be printed to std-out')
+    output.add_argument('-n', dest='outname', required=False,
+                        help='Define output file name - Mandatory is -o is provided: <outname>_<contig_id>_ORF_Comparison.csv')
+
+    misc = parser.add_argument_group('Misc')
+    misc.add_argument('-v', dest='verbose', default='False', type=eval, choices=[True, False],
+                        help='Default - False: Print out runtime status')
+    options = parser.parse_args()
+    comparator(options)
+
+if __name__ == "__main__":
+    main()
+    print("Complete")
+