#Author, date: Auden Cote-L'Heureux, July 17 2023 #Motivation: Run Guidance on multiple files #Intent: To be used in place of PhyloToL part 2 if there is a good reason not to use the guidance-only mode of PhyloToL #Dependencies: Python3, BioPython #Inputs: A folder of unaligned fasta files #Outputs: Folder for each input file with selected Guidance output files, renamed with the original file name at the front #Example: python3 GuidanceWrapper_v2.1.py -i FastaFiles #Dependencies import os, sys import argparse from Bio import SeqIO #Reading arguments parser = argparse.ArgumentParser( prog = 'Guidance Wrapper v2.1', description = "Updated July 21, 2023 by Auden Cote-L'Heureux" ) parser.add_argument('--input', '-i', required = True, type = str, help = 'Path to folder of unaligned amino acid fasta files to align. File extensions must be fasta, fa, fas, or faa. Try using the absolute rather than relative path if working on the Grid and having trouble') parser.add_argument('--output', '-o', default = '.', type = str, help = 'Path to folder where a folder named "GuidanceOutput" will be created') parser.add_argument('--codon', action = 'store_true', help = 'Run on nucleotide files by translating codons') parser.add_argument('--iterations', '-n', default = 5, type = int, help = 'Number of Guidance iterations (default = 5)') parser.add_argument('--guidance_path', '-p', default = 'guidance.v2.02', type = str, help = 'Path to the guidance_v2.02 folder') parser.add_argument('--seq_cutoff', '-s', default = 0.3, type = float, help = 'Taxa are removed if their score is below this cutoff') parser.add_argument('--col_cutoff', '-c', default = 0.0, type = float, help = 'Columns are removed if their score is below this cutoff') parser.add_argument('--res_cutoff', '-r', default = 0.0, type = float, help = 'During guidance, residues are removed if their score is below this cutoff') parser.add_argument('--force', '-f', action = 'store_true', help = 'Delete existing output folder at given output path') parser.add_argument('--keep_temp', '-k', action = 'store_true', help = 'Keep all Guidance intermediate files (beware, some can be very large)') parser.add_argument('--guidance_threads', '-t', default = 20, type = int, help = 'Number of threads to allocate to Guidance') args = parser.parse_args() #Creating an output folder if not os.path.isdir(args.output + '/GuidanceOutput'): os.mkdir(args.output + '/GuidanceOutput') elif not args.force: print('\nERROR: An output folder already exists at the given path. Delete it or run in --force mode\n') quit() #For each file in the input folder for file in os.listdir(args.input): #If it has an appropriate extension for an amino acid fasta file if file.split('.')[-1] in ('fasta', 'fa', 'fas', 'faa'): #Create the output folder tax_guidance_outdir = args.output + '/GuidanceOutput/' + file.split('.')[0] if(not os.path.isdir(tax_guidance_outdir)): os.mkdir(tax_guidance_outdir) #Read in the sequence data and replace any * and U characters with 'X' recs = [r for r in SeqIO.parse(args.input + '/' + file, 'fasta')] with open(args.input + '/' + file, 'w') as o: for rec in recs: o.write('>' + rec.description + '\n' + str(rec.seq).replace('*', 'X').replace('U', 'X') + '\n\n') fail = False #For each Guidance iteration (default = 5) for i in range(args.iterations): #Get the number sequences in the unaligned file n_recs = len([r for r in SeqIO.parse(args.input + '/' + file, 'fasta')]) #Cancel the Guidance run if there are fewer than 4 sequences if n_recs < 4: print('\nGene famiily ' + file.split('.')[0].split('_preguidance')[0] + ' contains fewer than 4 sequences after ' + str(i) + ' Guidance iterations, therefore no alignment will be produced for this gene family\n') os.system('rm -rf ' + tax_guidance_outdir) if i == 0: fail = True break #MAFFT parameters depend on sequence count if n_recs < 200: mafft_alg = 'genafpair' else: mafft_alg = 'auto' if args.codon: seqtype = 'codon' else: seqtype = 'aa' #Run Guidance os.system(args.guidance_path + '/www/Guidance/guidance.pl --seqFile ' + args.input + '/' + file + ' --msaProgram MAFFT --seqType ' + seqtype + ' --outDir ' + tax_guidance_outdir + ' --seqCutoff ' + str(args.seq_cutoff) + ' --colCutoff ' + str(args.col_cutoff) + " --outOrder as_input --bootstraps 10 --MSA_Param '\\--" + mafft_alg + " --maxiterate 1000 --thread " + str(args.guidance_threads) + " --bl 62 --anysymbol' > " + tax_guidance_outdir + '/log.txt') #If it ran successfully if os.path.isfile(tax_guidance_outdir + '/MSA.MAFFT.Guidance2_res_pair_seq.scr_with_Names'): sep = ' ' if '\t' in open(tax_guidance_outdir + '/MSA.MAFFT.Guidance2_res_pair_seq.scr_with_Names').readlines()[1]: sep = '\t' #Create a record of sequences below the sequence score cutoff seqs_below = len([line for line in open(tax_guidance_outdir + '/MSA.MAFFT.Guidance2_res_pair_seq.scr_with_Names').readlines()[1:-1] if float(line.split(sep)[-1]) < args.seq_cutoff]) #If the number of remaining sequences is less than 4, then stop iterating if n_recs - seqs_below < 4: print('\nGene famiily ' + file.split('.')[0].split('_preguidance')[0] + ' contains fewer than 4 sequences after ' + str(i + 1) + ' Guidance iterations, therefore no alignment will be produced for this gene family.\n') os.system('rm -rf ' + tax_guidance_outdir) break #If there are no sequences below the cutoff or the number of Guidance iterations has been reached, then stop iterating if seqs_below == 0 or i == args.iterations - 1: print('\nGuidance complete after ' + str(i + 1) + ' iterations for gene family ' + file.split('.')[0].split('_preguidance')[0] + '\n') break #If another iteration is needed, then replace the original input file with the sequence-filtered run to input to Guidance for the next iteration os.system('cp ' + tax_guidance_outdir + '/Seqs.Orig.fas.FIXED.Without_low_SP_Seq.With_Names ' + args.input + '/' + file) #If another iteration is needed, remove the Guidance output from the last iteration os.system('rm -r ' + tax_guidance_outdir + '/*') else: fail = True break #If Guidance ran successfully if not fail: #Create a record of sequences above the sequence score cutoff seqs2keep = [rec.description for rec in SeqIO.parse(tax_guidance_outdir + '/Seqs.Orig.fas.FIXED.Without_low_SP_Seq.With_Names', 'fasta')] orig_seqs = [rec.description for rec in SeqIO.parse(tax_guidance_outdir + '/MSA.MAFFT.aln.With_Names', 'fasta')] #Put this sequence information in a dictionary and write out the surviving sequences to an unaligned .fasta file running_aln = { rec.description : str(rec.seq) for rec in SeqIO.parse(tax_guidance_outdir + '/MSA.MAFFT.aln.With_Names', 'fasta') if rec.description in seqs2keep } with open(tax_guidance_outdir + '/postGuidance_preMAFFT_unaligned.fasta', 'w') as o: for seq in running_aln: o.write('>' + seq + '\n' + str(running_aln[seq]).replace('-', '') + '\n\n') #Align the file with MAFFT os.system('mafft ' + tax_guidance_outdir + '/postGuidance_preMAFFT_unaligned.fasta > ' + tax_guidance_outdir + '/' + file.split('.')[0].split('_preguidance')[0] + '.postGuidance_MAFFT_realigned.fasta') #Read in the MAFFT alignment running_aln = { rec.description : str(rec.seq) for rec in SeqIO.parse(tax_guidance_outdir + '/' + file.split('.')[0].split('_preguidance')[0] + '.postGuidance_MAFFT_realigned.fasta', 'fasta') } sep = ' ' if '\t' in open(tax_guidance_outdir + '/MSA.MAFFT.Guidance2_res_pair_seq.scr').readlines()[1]: sep = '\t' #Apply residue cutoff per site per sequence for site in [(int(line.split(sep)[1]), int(line.split(sep)[0]) - 1) for line in open(tax_guidance_outdir + '/MSA.MAFFT.Guidance2_res_pair_seq.scr').readlines()[1:-1] if float(line.split(sep)[-1].strip()) < args.res_cutoff]: if(orig_seqs[site[0]] in seqs2keep): running_aln[orig_seqs[site[0]]][site[1]] = 'X' sep = ' ' if '\t' in open(tax_guidance_outdir + '/MSA.MAFFT.Guidance2_res_pair_col.scr').readlines()[1]: sep = '\t' #Apply column cutoff per column cols2remove = [int(line.split(sep)[0]) - 1 for line in open(tax_guidance_outdir + '/MSA.MAFFT.Guidance2_res_pair_col.scr').readlines()[1:-1] if float(line.split(sep)[-1].strip()) < args.col_cutoff] for seq in running_aln: running_aln[seq] = ''.join([running_aln[seq][i] for i in range(len(running_aln[seq])) if i not in cols2remove]) #Write out the final filtered alignment with open(tax_guidance_outdir + '/' + file.split('.')[0].split('_preguidance')[0] + '.postGuidance_MAFFT_realigned.fasta', 'w') as o: for seq in running_aln: o.write('>' + seq + '\n' + str(running_aln[seq]) + '\n\n') #If the user hasn't specified to keep all intermediate files, then delete the extraneous ones and rename the rest per the input file name if not args.keep_temp: for gdir_file in os.listdir(tax_guidance_outdir): if gdir_file not in ('MSA.MAFFT.Guidance2_res_pair_seq.scr_with_Names', 'MSA.MAFFT.aln.With_Names', 'MSA.MAFFT.Guidance2_res_pair_col.scr', 'log', file.split('.')[0].split('_preguidance')[0] + '.postGuidance_MAFFT_realigned.fasta'): os.system('rm -r ' + tax_guidance_outdir + '/' + gdir_file) else: if gdir_file == 'MSA.MAFFT.aln.With_Names': os.system('mv ' + tax_guidance_outdir + '/' + gdir_file + ' ' + tax_guidance_outdir + '/' + file.split('.')[0].split('_preguidance')[0] + '_' + gdir_file + '.aln') else: os.system('mv ' + tax_guidance_outdir + '/' + gdir_file + ' ' + tax_guidance_outdir + '/' + file.split('.')[0].split('_preguidance')[0] + '_' + gdir_file)