Software_comparison.py

import pandas as pd
import os
import time
import glob

import Benchmarking_paper_only.FDR as FDR
import TPP_reusable.Confident_PTMs as Confident_PTMs
import TPP_reusable.Unique_PTMs as Unique_PTMs
import Benchmarking_paper_only.extract_scan as extract_scan
import TPP_reusable.convert_xml_sax as convert_xml_sax
import TPP_reusable.Comparison as Comparison
import TPP_reusable.Post_analysis as Post_analysis
import TPP_reusable.pAla as pAla
import sys
from params_software import *

start_time = time.time()
Ascore_cutoff = 0
sub = "FDR_" + str(FDR_cutoff) + "_PTM_score_" + str(Ascore_cutoff)
decoy_list=[]
for wd, software in zip(w, s):
    try:
        os.chdir(wd)
    except OSError:
        print("Can't find results directory, please check location: ")
        print(wd)
    if "PEAKS" in software:
        input = "DB search psm.csv"
        results_file = "DB search psm_edit.csv"
        # extract scan from peaks output
        if os.path.isfile(results_file):
            print("Scan already found")
        else:
            extract_scan.scan(search_files_location, input, results_file, wd)
            print("Scan found")
        search_software="PEAKS"
    elif "TPP" in software:
        # Run pepXML converter - if csv not in dir
        results_file = "interact.ptm.pep.csv"
        if os.path.isfile(results_file):
            print("XML converted file exists")
        else:
            xml = glob.glob("*.ptm.pep.xml")
        try:
            print("Converting XML file")
            convert_xml_sax.convert(xml[0])
            print("XML converted")
        except ValueError:
            raise ValueError("Please provide TPP '.ptm.pep.xml' results file")
        search_software = "TPP"
    elif "MaxQuant" in software:
        results_file = "evidence.txt"
        if not os.path.isfile(results_file):
            print("Please provide location for 'evidence.txt' MaxQuant results file")
        search_software = "MaxQuant"
    elif "PD" in software:
        text_file = glob.glob("*PSMs.txt")
        if len(text_file) != 1:
            raise ValueError('should be only one txt file in the PD directory')
        results_file = text_file[0]
        search_software = "PD"
    # calculate FDR
    if not os.path.exists(sub):
        os.mkdir(sub)
    FDR_output = sub + "/FDR_output.csv"
    if "pL" in software or "pL" in wd or "pLeu" in wd:
        search="STYL"
        decoy="pLeu"
    elif "pG" in software or "pG" in wd or "pGly" in wd:
        search="STYG"
        decoy="pGly"
    elif "pD" in software or "pD" in wd or "pAsp" in wd:
        search="STYD"
        decoy="pAsp"
    elif "pE" in software or "pE" in wd or "pGlu" in wd:
        search="STYE"
        decoy="pGlu"
    elif "pP" in software or "pP" in wd or "pPro" in wd:
        search="STYP"
        decoy="pPro"
    else:
        search="STYA"
        decoy="pAla"
    decoy_list.append(decoy)
    FDR.calculateFDR(results_file, search_software, FDR_output, PXD, wd, search)
    print("FDR done")
    print("--- %s seconds ---" % (time.time() - start_time))
    # filter for FDR (and A-score) cutoff
    confident_PTM_output = sub + "/PTM_confident.csv"
    Confident_PTMs.confident(FDR_output, database, FDR_cutoff, Ascore_cutoff, confident_PTM_output)
    print("Confident PTMs done")
    print("--- %s seconds ---" % (time.time() - start_time))
    # Collapse for best scoring for each peptide, protein site, mass shift on protein or no collapse
    unique_peptide = sub + "/Peptide_confident_PTM_unique.csv"
    unique_site = sub + "/Site_confident_PTM_unique.csv"
    unique_mass = sub + "/Peptide_mass_confident_PTM_unique.csv"
    non_collapse = sub + "/All_confident_PTM_no_collapse.csv"
    #If synthetic dataset, use alternative analysis (include pool etc.)
    if PXD=="PXD007058":
        sys.exit("Use Synthetics Analysis Pipeline")
    else:
        Unique_PTMs.unique(confident_PTM_output, unique_peptide, unique_site, unique_mass, non_collapse)
    print("Unique PTMs done")
    print(FDR_cutoff, software, "--- %s seconds ---" % (time.time() - start_time))

# Compare spectrum between different searches
files = []
for i in w:
    files.append(i + "/"+  sub + "/PTM_confident.csv")
Comparison.comparison(p, sub, files, s)
print("Spectrum Comparison Done --- %s seconds ---" % (time.time() - start_time))

file_input_list = ["All_confident_PTM_no_collapse.csv"]
files = []
for i in w:
    files.append(i + "/" + sub + "/")
software_list_edit = ""
for i in s:
    software_list_edit += i + "_"
os.chdir(p)
if not os.path.exists("Comparisons"):
    os.mkdir("Comparisons")
# load in list of hits where spectrum comparisons gave match between all searches
spectrum_match_list = []
spectrum = p + "\Spectrum_Comparisons/" + software_list_edit[
                                          :-1] + "/" + sub + "_spectrum_comparison_" + software_list_edit[
                                                                                       :-1] + ".csv"
df = pd.read_csv(spectrum, dtype=str)
for i in range(len(df)):
    if df.loc[i, 'Match'] == "TRUE":
        spectrum_match_list.append(df.loc[i, 'Spectrum'])

print("Starting FLR calculations --- %s seconds ---" % (time.time() - start_time))
#Post analysis - FLR calulations and plots
files = []
for i in w:
    for f in file_input_list:
        files.append(i + "/" + sub + "/" + f)
        
for file in files:
    Post_analysis.site_input = Post_analysis.site_based(file)
    Post_analysis.site_all_input=Post_analysis.site_based_all(file)
    output = file.replace(".csv", "_Site-based.csv")
    Post_analysis.spectrum_comparison(output, spectrum_match_list)
final_file_list = []
final_file_list_collapsed = []
final_file_list_filtered =[]
final_file_list_PTM=[]
for i,decoy in zip(w,decoy_list):
    Post_analysis.model_FLR(i+"/"+sub+"/"+"All_confident_PTM_no_collapse_Site-based_spectrum_match.csv")
    Post_analysis.model_FLR_new(i + "/" + sub + "/" + "All_confident_PTM_no_collapse_Site-based_spectrum_match.csv")
    Post_analysis.model_FLR_filter(i + "/" + sub + "/" + "All_confident_PTM_no_collapse_Site-based_spectrum_match.csv")
    pAla.calulate_decoy_FLR(i+"/"+sub+"/"+"All_confident_PTM_no_collapse_Site-based_spectrum_match_FLR.csv",decoy)
    pAla.calulate_decoy_FLR(i + "/" + sub + "/" + "All_confident_PTM_no_collapse_Site-based_spectrum_match_new_FLR_collapse.csv", decoy)
    pAla.calulate_decoy_FLR(i+"/"+sub+"/"+"All_confident_PTM_no_collapse_Site-based_spectrum_match_FLR_filtered.csv",decoy)
    final_file_list.append(i+"/"+sub+"/"+"All_confident_PTM_no_collapse_Site-based_spectrum_match_FLR_"+decoy+".csv")    
    final_file_list_collapsed.append(i+"/"+sub+"/"+"All_confident_PTM_no_collapse_Site-based_spectrum_match_new_FLR_collapse_"+decoy+".csv")
    final_file_list_filtered.append(i+"/"+sub+"/"+"All_confident_PTM_no_collapse_Site-based_spectrum_match_FLR_filtered_"+decoy+".csv")
    Post_analysis.PTM_sort_FLR(i+"/"+sub+"/"+"All_confident_PTM_no_collapse_Site-based_spectrum_match.csv")
    pAla.calulate_decoy_FLR(i + "/" + sub + "/" + "All_confident_PTM_no_collapse_Site-based_spectrum_match_FLR_PTM_sort.csv",decoy)
    final_file_list_PTM.append(i+"/"+sub+"/"+"All_confident_PTM_no_collapse_Site-based_spectrum_match_FLR_PTM_sort_"+decoy+".csv")

Post_analysis.plot_FLR_comparisons(sub, final_file_list, s, p)
Post_analysis.plot_FLR_comparisons(sub, final_file_list_collapsed, s, p)
Post_analysis.plot_FLR_comparisons_PTM_prob(sub,final_file_list_PTM,s,p)
Post_analysis.plot_FLR_comparisons(sub,final_file_list_filtered,s,p)

final_file_list = []
final_file_list_collapse=[]
for i,decoy in zip(w,decoy_list):
    Post_analysis.model_FLR(i+"/"+sub+"/"+"Site_confident_PTM_unique.csv")
    Post_analysis.model_FLR_new(i+"/"+sub+"/"+"Site_confident_PTM_unique.csv")
    pAla.calulate_decoy_FLR(i+"/"+sub+"/"+"Site_confident_PTM_unique_FLR.csv",decoy)
    pAla.calulate_decoy_FLR(i+"/"+sub+"/"+"Site_confident_PTM_unique_new_FLR_collapse.csv",decoy)
    final_file_list.append(i+"/"+sub+"/"+"Site_confident_PTM_unique_FLR_"+decoy+".csv")
    final_file_list_collapse.append(i + "/" + sub + "/" + "Site_confident_PTM_unique_new_FLR_collapse_" + decoy + ".csv")
Post_analysis.plot_FLR_comparisons(sub, final_file_list, s, p)
Post_analysis.plot_FLR_comparisons(sub, final_file_list_collapse, s, p)
print("FLR calculations done --- %s seconds ---" % (time.time() - start_time))