class TestConverter(unittest.TestCase): def setUp(self): self.converter = Converter() self.example = Example() self.converter.gff3parser = Mock_gff3_parser self.converter._print_rntptt_title = Mock_func().print_rntptt_title self.converter.tsspredator = Mock_TSSPredatorReader() self.converter._read_file = Mock_func().mock_read_file self.gff_file = self.example.gff_file self.ptt_out = self.example.ptt_out self.rnt_out = self.example.rnt_out self.srna_out = self.example.srna_out self.embl_file = self.example.embl_file self.embl_out = self.example.embl_out self.multi_embl = self.example.multi_embl self.gff_out = self.example.gff_out self.mastertable = self.example.mastertable self.tss_file = self.example.tss_file self.fasta_file = self.example.fasta_file self.transterm = self.example.transterm self.term_file = self.example.term_file self.circ_file = self.example.circrna_table self.circ_all = self.example.circrna_all self.circ_best = self.example.circrna_best self.test_folder = "test_folder" self.mock_args = MockClass() if (not os.path.exists(self.test_folder)): os.mkdir(self.test_folder) def tearDown(self): if os.path.exists(self.test_folder): shutil.rmtree(self.test_folder) def test_print_rntptt_file(self): cdss = [] genes = [] rnas = [] gff_dict = Example().gff_dict for gff in gff_dict: if gff["feature"] == "gene": genes.append(self.converter.gff3parser.entries(self, gff)) elif gff["feature"] == "CDS": cdss.append(self.converter.gff3parser.entries(self, gff)) elif gff["feature"] == "tRNA": rnas.append(self.converter.gff3parser.entries(self, gff)) out_p = StringIO() out_r = StringIO() self.converter._print_rntptt_file(out_p, cdss, genes) self.converter._print_rntptt_file(out_r, rnas, genes) self.assertEqual(out_p.getvalue().split("\n")[:-1], self.example.ptt_out_list) self.assertEqual(out_r.getvalue().split("\n")[:-1], self.example.rnt_out_list) out_p.close() out_r.close() def test_srna2pttrnt(self): srna_input_file = os.path.join(self.test_folder, "srna.gff") srna_output_file = os.path.join(self.test_folder, "srna.out") with open(srna_input_file, "w") as fh: fh.write(self.gff_file) srnas = [] self.converter._srna2rntptt(srna_input_file, srna_output_file, srnas, 1234567) datas = import_data(srna_output_file) self.assertEqual(set(datas), set(self.srna_out.split("\n"))) def test_multi_embl_pos(self): embls = [] for line in self.embl_file.split("\n"): datas = self.converter._multi_embl_pos(line.strip()) if datas != "Wrong": embls.append(datas) for index in range(0, 7): self.assertDictEqual(embls[index], self.embl_out[index]) for index in range(0, 2): self.assertDictEqual(embls[-1]["pos"][index], self.multi_embl[index]) def test_parser_embl_data(self): embl_file = os.path.join(self.test_folder, "test.embl") embl_out = os.path.join(self.test_folder, "test.embl_out") out = StringIO() with open(embl_file, "w") as eh: for line in self.embl_file.split("\n"): eh.write(line + "\n") info = self.converter._parser_embl_data(embl_file, out) datas = out.getvalue().split("\n") self.assertEqual(set(datas[:-1]), set(self.gff_out.split("\n"))) self.assertEqual(info[0], "NC_007795.1") for index in range(0, 2): self.assertDictEqual(info[1]["pos"][index], self.multi_embl[index]) out.close() def test_multi_tss_class(self): nums = {"tss": 0, "tss_uni": 0, "class": 1} utrs = {"total": [], "pri": [], "sec": []} tss_features = {"tss_types": [], "locus_tags": [], "utr_lengths": []} tss_index = defaultdict(lambda: 0) master_file = os.path.join(self.test_folder, "test.tsv") fh = StringIO(self.mastertable) for tss in self.converter.tsspredator.entries(fh): self.converter._multi_tss_class(tss, tss_index, tss_features, nums, utrs) fh.close() self.assertDictEqual(nums, {'tss_uni': 0, 'class': 5, 'tss': 2}) def test_convert_mastertable2gff(self): master_file = os.path.join(self.test_folder, "test.tsv") with open(master_file, "w") as th: th.write(self.mastertable) out_gff = os.path.join(self.test_folder, "test.tsv_out") self.converter.convert_mastertable2gff(master_file, "ANNOgesic", "TSS", "aaa", out_gff) datas = import_data(out_gff) self.assertEqual(set(datas), set(self.tss_file.split("\n"))) def test_convert_gff2rntptt(self): srna_input_file = os.path.join(self.test_folder, "srna.gff") srna_output_file = os.path.join(self.test_folder, "srna.out") gff_file = os.path.join(self.test_folder, "test.gff") rnt_file = os.path.join(self.test_folder, "test.rnt") ptt_file = os.path.join(self.test_folder, "test.ptt") fasta_file = os.path.join(self.test_folder, "test.fa") with open(srna_input_file, "w") as fh: fh.write(self.gff_file) with open(gff_file, "w") as fh: fh.write(self.gff_file) with open(fasta_file, "w") as fh: fh.write(self.fasta_file) self.converter.convert_gff2rntptt(gff_file, fasta_file, ptt_file, rnt_file, srna_input_file, srna_output_file) self.assertTrue(srna_output_file) self.assertTrue(rnt_file) self.assertTrue(ptt_file) def test_convert_embl2gff(self): embl_file = os.path.join(self.test_folder, "test.embl") gff_file = os.path.join(self.test_folder, "test.embl_out") with open(embl_file, "w") as eh: for line in self.embl_file.split("\n"): eh.write(line + "\n") self.converter.convert_embl2gff(embl_file, gff_file) datas = import_data(gff_file) self.assertEqual(set(datas[1:-2]), set(self.gff_out.split("\n"))) def test_convert_transtermhp2gff(self): transterm_file = os.path.join(self.test_folder, "test_best_terminator_after_gene.bag") gff_file = os.path.join(self.test_folder, "transterm.gff") with open(transterm_file, "w") as th: th.write(self.transterm) self.converter.convert_transtermhp2gff(transterm_file, gff_file) datas = import_data(gff_file) self.assertEqual(set(datas), set(self.term_file.split("\n"))) def get_info(datas): f_datas = [] for data in datas: if not data.startswith("#"): f_datas.append("\t".join(data.split("\t")[:8])) return f_datas def test_convert_circ2gff(self): circ_file = os.path.join(self.test_folder, "circ.csv") out_all = os.path.join(self.test_folder, "all.gff") out_filter = os.path.join(self.test_folder, "best.gff") with open(circ_file, "w") as ch: ch.write(self.circ_file) args = self.mock_args.mock() args.start_ratio = 0.5 args.end_ratio = 0.5 args.support = 5 self.converter.convert_circ2gff(circ_file, args, out_all, out_filter) datas = import_data(out_all) f_datas = [] for data in datas: if not data.startswith("#"): f_datas.append("\t".join(data.split("\t")[:8])) c_datas = [] for data in self.circ_all.split("\n"): if not data.startswith("#"): c_datas.append("\t".join(data.split("\t")[:8])) self.assertListEqual(f_datas, c_datas) datas = import_data(out_filter) f_datas = [] for data in datas: if not data.startswith("#"): f_datas.append("\t".join(data.split("\t")[:8])) c_datas = [] for data in self.circ_best.split("\n"): if not data.startswith("#"): c_datas.append("\t".join(data.split("\t")[:8])) self.assertListEqual(f_datas, c_datas)
class TSSpredator(object): def __init__(self, args_tss): self.multiparser = Multiparser() self.helper = Helper() self.converter = Converter() self.master = os.path.join(args_tss.out_folder, "MasterTables") self.tmps = { "tss": "tmp_TSS", "ta_tss": "tmp_ta_tss", "tss_ta": "tmp_tss", "tmp": "tmp" } if args_tss.ta_files is not None: self.tmps["ta"] = os.path.join(args_tss.ta_files, "tmp") else: self.tmps["ta"] = None self.gff_path = os.path.join(args_tss.gffs, "tmp") self.wig_path = os.path.join(args_tss.wig_folder, "tmp") self.fasta_path = os.path.join(args_tss.fastas, "tmp") self.stat_outfolder = os.path.join(args_tss.out_folder, "statistics") self.gff_outfolder = os.path.join(args_tss.out_folder, "gffs") def _assign_dict(self, lib_datas): return { "wig": lib_datas[0], "tex": lib_datas[1], "condition": int(lib_datas[2]), "replicate": lib_datas[3], "strand": lib_datas[4] } def _print_lib(self, lib_num, lib_list, out, wig_folder, prefix, rep_set): for num_id in range(1, lib_num + 1): cond_list = [] for lib in lib_list: if num_id == lib["condition"]: cond_list.append(lib) cond_sort_list = sorted(cond_list, key=lambda k: k['replicate']) reps = [] for cond in cond_sort_list: out.write("{0}_{1}{2} = {3}\n".format( prefix, cond["condition"], cond["replicate"], os.path.join(wig_folder, cond["wig"]))) reps.append(cond["replicate"]) for rep in sorted(rep_set): if rep not in reps: out.write("{0}_{1}{2} = \n".format(prefix, cond["condition"], rep)) def _start_to_run(self, tsspredator_path, config_file, out_path, prefix): print("Running TSSpredator for " + prefix) out = open(os.path.join(out_path, "log.txt"), "w") err = open(os.path.join(out_path, "err.txt"), "w") call(["java", "-jar", tsspredator_path, config_file], stdout=out, stderr=err) out.close() err.close() def _import_lib(self, libs, wig_folder, project_strain_name, out, gff, program, fasta): lib_dict = {"fp": [], "fm": [], "nm": [], "np": []} lib_num = 0 rep_set = set() list_num_id = [] print("Runniun {0} now...".format(program)) for lib in libs: lib_datas = lib.split(":") if not lib_datas[0].endswith(".wig"): print("Error:Exist a not proper wig files!!") sys.exit() for wig in os.listdir(wig_folder): filename = wig.split("_STRAIN_") if (filename[0] == lib_datas[0][:-4]) and (filename[1][:-4] == project_strain_name): lib_datas[0] = wig if int(lib_datas[2]) > lib_num: lib_num = int(lib_datas[2]) if lib_datas[3] not in rep_set: rep_set.add(lib_datas[3]) if (lib_datas[1] == "tex") and (lib_datas[4] == "+"): lib_dict["fp"].append(self._assign_dict(lib_datas)) elif (lib_datas[1] == "tex") and (lib_datas[4] == "-"): lib_dict["fm"].append(self._assign_dict(lib_datas)) elif (lib_datas[1] == "notex") and (lib_datas[4] == "+"): lib_dict["np"].append(self._assign_dict(lib_datas)) elif (lib_datas[1] == "notex") and (lib_datas[4] == "-"): lib_dict["nm"].append(self._assign_dict(lib_datas)) for num_id in range(1, lib_num + 1): out.write("annotation_{0} = {1}\n".format(num_id, gff)) if program.lower() == "tss": self._print_lib(lib_num, lib_dict["fm"], out, wig_folder, "fivePrimeMinus", rep_set) self._print_lib(lib_num, lib_dict["fp"], out, wig_folder, "fivePrimePlus", rep_set) elif program.lower() == "processing_site": self._print_lib(lib_num, lib_dict["nm"], out, wig_folder, "fivePrimeMinus", rep_set) self._print_lib(lib_num, lib_dict["np"], out, wig_folder, "fivePrimePlus", rep_set) else: print("Error: Wrong program name!!!") sys.exit() for num_id in range(1, lib_num + 1): out.write("genome_{0} = {1}\n".format(num_id, fasta)) for num_id in range(1, lib_num + 1): list_num_id.append(str(num_id)) return lib_num, num_id, rep_set, lib_dict, list_num_id def _print_repmatch(self, args_tss, out): '''check replicate match''' if "all" in args_tss.repmatch: match = args_tss.repmatch.split("_")[-1] out.write("minNumRepMatches = {0}\n".format(match)) else: nums = {} matchs = {} for match in args_tss.repmatch.split(","): lib = match.split("_")[0] rep = match.split("_")[-1] matchs[lib] = rep if rep not in nums.keys(): nums[rep] = 1 else: nums[rep] += 1 for rep, num in nums.items(): if num == max(nums.values()): out.write("minNumRepMatches = {0}\n".format(rep)) max_rep = rep break for lib, rep in matchs.items(): if rep != max_rep: out.write("minNumRepMatches_{0} = {1}\n".format(lib, rep)) def _gen_config(self, project_strain_name, args_tss, gff, wig_folder, fasta, config_file): '''generation of config files''' master_folder = "MasterTable_" + project_strain_name out_path = os.path.join(self.master, master_folder) self.helper.check_make_folder(out_path) out = open(config_file, "w") out.write("TSSinClusterSelectionMethod = HIGHEST\n") out.write("allowedCompareShift = 1\n") out.write("allowedRepCompareShift = 1\n") lib_num, num_id, rep_set, lib_dict, list_num_id = \ self._import_lib(args_tss.libs, wig_folder, project_strain_name, out, gff, args_tss.program, fasta) out.write("idList = ") out.write(",".join(list_num_id) + "\n") out.write("maxASutrLength = 100\n") out.write("maxGapLengthInGene = 500\n") out.write("maxNormalTo5primeFactor = {0}\n".format( args_tss.processing_factor)) out.write("maxTSSinClusterDistance = {0}\n".format(args_tss.cluster + 1)) out.write("maxUTRlength = {0}\n".format(args_tss.utr_length)) out.write("min5primeToNormalFactor = {0}\n".format( args_tss.enrichment_factor)) out.write("minCliffFactor = {0}\n".format(args_tss.factor)) out.write("minCliffFactorDiscount = {0}\n".format( args_tss.factor_reduction)) out.write("minCliffHeight = {0}\n".format(args_tss.height)) out.write("minCliffHeightDiscount = {0}\n".format( args_tss.height_reduction)) out.write("minNormalHeight = {0}\n".format(args_tss.base_height)) self._print_repmatch(args_tss, out) out.write("minPlateauLength = 0\n") out.write("mode = cond\n") out.write("normPercentile = 0.9\n") if args_tss.program.lower() == "tss": self._print_lib(lib_num, lib_dict["nm"], out, wig_folder, "normalMinus", rep_set) self._print_lib(lib_num, lib_dict["np"], out, wig_folder, "normalPlus", rep_set) else: self._print_lib(lib_num, lib_dict["fm"], out, wig_folder, "normalMinus", rep_set) self._print_lib(lib_num, lib_dict["fp"], out, wig_folder, "normalPlus", rep_set) out.write("numReplicates = {0}\n".format(len(rep_set))) out.write("numberOfDatasets = {0}\n".format(lib_num)) out.write("outputDirectory = {0}\n".format(out_path)) for prefix_id in range(len(args_tss.output_prefixs)): out.write("outputPrefix_{0} = {1}\n".format( prefix_id + 1, args_tss.output_prefixs[prefix_id])) out.write("projectName = {0}\n".format(project_strain_name)) out.write("superGraphCompatibility = igb\n") out.write("texNormPercentile = 0.5\n") out.write("writeGraphs = 0\n") out.write("writeNocornacFiles = 0\n") out.close() def _convert_gff(self, prefixs, args_tss): for prefix in prefixs: out_file = os.path.join( self.gff_outfolder, "_".join([prefix, args_tss.program]) + ".gff") gff_f = open(out_file, "w") out_path = os.path.join(self.master, "_".join(["MasterTable", prefix])) if "MasterTable.tsv" not in os.listdir(out_path): print("Error:there is not MasterTable file in {0}".format( out_path)) print("Please check configuration file.") else: if args_tss.program.lower() == "processing": feature = "processing_site" elif args_tss.program.lower() == "tss": feature = "TSS" self.converter.convert_mastertable2gff( os.path.join(out_path, "MasterTable.tsv"), "ANNOgesic", feature, prefix, out_file) gff_f.close() def _merge_manual(self, tsss, args_tss): '''if manual detected TSS is provided, it can merge manual detected TSS and TSSpredator predicted TSS''' self.helper.check_make_folder( os.path.join(os.getcwd(), self.tmps["tss"])) for tss in tsss: for gff in os.listdir(args_tss.gffs): if (gff[:-4] == tss) and (".gff" in gff): break filename = "_".join([tss, args_tss.program]) + ".gff" predict = os.path.join(self.gff_outfolder, filename) print("Running merge and classify manual ....") stat_file = "stat_compare_TSSpredator_manual_{0}.csv".format(tss) merge_manual_predict_tss(predict, stat_file, os.path.join(self.tmps["tss"], filename), os.path.join(args_tss.gffs, gff), args_tss) shutil.move( stat_file, os.path.join(args_tss.out_folder, "statistics", tss, stat_file)) self.helper.move_all_content(self.tmps["tss"], self.gff_outfolder, [".gff"]) shutil.rmtree(self.tmps["tss"]) def _validate(self, tsss, args_tss): '''validate TSS with genome annotation''' print("Running validation of annotation....") for tss in tsss: for gff in os.listdir(args_tss.gffs): if (gff[:-4] == tss) and (".gff" in gff): break stat_file = os.path.join(self.stat_outfolder, tss, "".join(["stat_gene_vali_", tss, ".csv"])) out_cds_file = os.path.join(args_tss.out_folder, "tmp.gff") if args_tss.program.lower() == "tss": compare_file = os.path.join(self.gff_outfolder, "_".join([tss, "TSS.gff"])) elif args_tss.program.lower() == "processing": compare_file = os.path.join(self.gff_outfolder, "_".join([tss, "processing.gff"])) validate_gff(compare_file, os.path.join(args_tss.gffs, gff), stat_file, out_cds_file, args_tss.utr_length, args_tss.program.lower()) shutil.move(out_cds_file, os.path.join(args_tss.gffs, gff)) def _compare_ta(self, tsss, args_tss): '''compare TSS with transcript''' detect = False print("Running compare transcript assembly and TSS ...") self.multiparser.parser_gff(args_tss.ta_files, "transcript") self.multiparser.combine_gff(args_tss.gffs, self.tmps["ta"], None, "transcript") for tss in tsss: stat_out = os.path.join( self.stat_outfolder, tss, "".join(["stat_compare_TSS_transcript_", tss, ".csv"])) for ta in os.listdir(self.tmps["ta"]): filename = ta.split("_transcript") if (filename[0] == tss) and (filename[1] == ".gff"): detect = True break compare_file = os.path.join(self.gff_outfolder, "_".join([tss, "TSS.gff"])) if detect: stat_ta_tss(os.path.join(self.tmps["ta"], ta), compare_file, stat_out, self.tmps["ta_tss"], self.tmps["tss_ta"], args_tss.fuzzy) self.helper.sort_gff(self.tmps["tss_ta"], compare_file) self.helper.sort_gff(self.tmps["ta_tss"], os.path.join(args_tss.ta_files, ta)) os.remove(self.tmps["tss_ta"]) os.remove(self.tmps["ta_tss"]) detect = False def _stat_tss(self, tsss, feature): print("Running statistaics.....") for tss in tsss: compare_file = os.path.join(self.gff_outfolder, "_".join([tss, feature]) + ".gff") stat_tsspredator( compare_file, feature, os.path.join( self.stat_outfolder, tss, "_".join(["stat", feature, "class", tss]) + ".csv"), os.path.join(self.stat_outfolder, tss, "_".join(["stat", feature, "libs", tss]) + ".csv")) self.helper.move_all_content( os.getcwd(), os.path.join(self.stat_outfolder, tss), ["_class", ".png"]) if os.path.exists( os.path.join(self.stat_outfolder, "TSSstatistics.tsv")): shutil.move( os.path.join(self.stat_outfolder, "TSSstatistics.tsv"), os.path.join(self.stat_outfolder, tss, "TSSstatistics.tsv")) plot_venn(compare_file, feature) self.helper.move_all_content( os.getcwd(), os.path.join(self.stat_outfolder, tss), ["_venn", ".png"]) def _set_gen_config(self, args_tss, input_folder): prefixs = [] detect = False for fasta in os.listdir(self.fasta_path): for gff in os.listdir(self.gff_path): if fasta[:-3] == gff[:-4]: prefix = fasta[:-3] for wig in os.listdir(self.wig_path): filename = wig.split("_STRAIN_") if filename[1][:-4] == prefix: detect = True break if detect: prefixs.append(prefix) config = os.path.join( input_folder, "_".join(["config", prefix]) + ".ini") self._gen_config(prefix, args_tss, os.path.join(self.gff_path, gff), self.wig_path, os.path.join(self.fasta_path, fasta), config) return prefixs def _merge_wigs(self, wig_folder, prefix, libs): self.helper.check_make_folder( os.path.join(os.getcwd(), self.tmps["tmp"])) for wig_file in os.listdir(wig_folder): for lib in libs: info = lib.split(":") if (info[0][:-4] in wig_file) and (info[-1] == "+") and ( prefix in wig_file) and (os.path.isfile( os.path.join(wig_folder, wig_file))): Helper().merge_file( os.path.join(wig_folder, wig_file), os.path.join("tmp", "merge_forward.wig")) if (info[0][:-4] in wig_file) and (info[-1] == "-") and ( prefix in wig_file) and (os.path.isfile( os.path.join(wig_folder, wig_file))): Helper().merge_file( os.path.join(wig_folder, wig_file), os.path.join("tmp", "merge_reverse.wig")) def _check_orphan(self, prefixs, wig_folder, args_tss): '''if genome has no locus tag, it can use for classify the TSS''' for prefix in prefixs: self._merge_wigs(wig_folder, prefix, args_tss.libs) tmp_tss = os.path.join( self.tmps["tmp"], "_".join([prefix, args_tss.program + ".gff"])) pre_tss = os.path.join( self.gff_outfolder, "_".join([prefix, args_tss.program + ".gff"])) check_orphan(pre_tss, os.path.join(args_tss.gffs, prefix + ".gff"), "tmp/merge_forward.wig", "tmp/merge_reverse.wig", tmp_tss) shutil.move(tmp_tss, pre_tss) shutil.rmtree("tmp") def _remove_files(self, args_tss): print("Remove temperary files and folders...") self.helper.remove_tmp(args_tss.fastas) self.helper.remove_tmp(args_tss.gffs) self.helper.remove_tmp(args_tss.wig_folder) self.helper.remove_tmp(args_tss.ta_files) if "merge_forward.wig" in os.listdir(os.getcwd()): os.remove("merge_forward.wig") if "merge_reverse.wig" in os.listdir(os.getcwd()): os.remove("merge_reverse.wig") def _deal_with_overlap(self, out_folder, args_tss): '''deal with the situation that TSS and processing site at the same position''' if args_tss.overlap_feature.lower() == "both": pass else: print("Comparing TSS and Processing site...") if args_tss.program.lower() == "tss": for tss in os.listdir(out_folder): if tss.endswith("_TSS.gff"): ref = self.helper.get_correct_file( args_tss.references, "_processing.gff", tss.replace("_TSS.gff", ""), None, None) filter_tss_pro(os.path.join(out_folder, tss), ref, args_tss.overlap_feature, args_tss.cluster) elif args_tss.program.lower() == "processing_site": for tss in os.listdir(out_folder): if tss.endswith("_processing.gff"): ref = self.helper.get_correct_file( args_tss.references, "_TSS.gff", tss.replace("_processing.gff", ""), None, None) filter_tss_pro(os.path.join(out_folder, tss), ref, args_tss.overlap_feature, args_tss.cluster) def _low_expression(self, args_tss, gff_folder): '''deal with the low expressed TSS''' prefix = None self._merge_wigs(args_tss.wig_folder, "wig", args_tss.libs) for gff in os.listdir(gff_folder): if (args_tss.program.lower() == "tss") and (gff.endswith("_TSS.gff")): prefix = gff.replace("_TSS.gff", "") elif (args_tss.program.lower() == "processing") and (gff.endswith("_processing.gff")): prefix = gff.replace("_processing.gff", "") if prefix: out = open( os.path.join( self.stat_outfolder, prefix, "_".join(["stat", prefix, "low_expression_cutoff.csv"])), "w") out.write("\t".join(["strain", "cutoff_coverage"]) + "\n") cutoff = filter_low_expression( os.path.join(gff_folder, gff), args_tss, "tmp/merge_forward.wig", "tmp/merge_reverse.wig", "tmp/without_low_expression.gff") out.write("\t".join([prefix, str(cutoff)]) + "\n") os.remove(os.path.join(gff_folder, gff)) shutil.move("tmp/without_low_expression.gff", os.path.join(gff_folder, gff)) prefix = None out.close() def run_tsspredator(self, args_tss): input_folder = os.path.join(args_tss.out_folder, "configs") for gff in os.listdir(args_tss.gffs): if gff.endswith(".gff"): self.helper.check_uni_attributes( os.path.join(args_tss.gffs, gff)) self.helper.check_make_folder(self.gff_outfolder) self.multiparser.parser_fasta(args_tss.fastas) self.multiparser.parser_gff(args_tss.gffs, None) self.multiparser.parser_wig(args_tss.wig_folder) prefixs = self._set_gen_config(args_tss, input_folder) for prefix in prefixs: out_path = os.path.join(self.master, "_".join(["MasterTable", prefix])) config_file = os.path.join(input_folder, "_".join(["config", prefix]) + ".ini") self._start_to_run(args_tss.tsspredator_path, config_file, out_path, prefix) if os.path.exists(os.path.join(out_path, "TSSstatistics.tsv")): shutil.move( os.path.join(out_path, "TSSstatistics.tsv"), os.path.join(self.stat_outfolder, "TSSstatistics.tsv")) if args_tss.program.lower() == "processing_site": args_tss.program = "processing" self._convert_gff(prefixs, args_tss) if args_tss.check_orphan: print("checking the orphan TSS...") self._check_orphan(prefixs, os.path.join(args_tss.wig_folder, "tmp"), args_tss) self.multiparser.combine_gff(args_tss.gffs, self.gff_outfolder, None, args_tss.program) datas = [] for gff in os.listdir(self.gff_outfolder): if gff.endswith(".gff"): gff_folder = gff.replace( "".join(["_", args_tss.program, ".gff"]), "") self.helper.check_make_folder( os.path.join(self.stat_outfolder, gff_folder)) datas.append(gff_folder) if args_tss.remove_low_expression is not None: self._low_expression(args_tss, self.gff_outfolder) if args_tss.manual is not None: self.multiparser.combine_wig(args_tss.gffs, self.wig_path, None, args_tss.libs) self._merge_manual(datas, args_tss) self._deal_with_overlap(self.gff_outfolder, args_tss) if args_tss.stat: self._stat_tss(datas, args_tss.program) if args_tss.validate: self._validate(datas, args_tss) if args_tss.ta_files is not None: self._compare_ta(datas, args_tss) self._remove_files(args_tss)
class TSSpredator(object): def __init__(self, args_tss): self.multiparser = Multiparser() self.helper = Helper() self.converter = Converter() self.master = os.path.join(args_tss.out_folder, "MasterTables") self.tmps = {"tss": "tmp_TSS", "ta_tss": "tmp_ta_tss", "tss_ta": "tmp_tss", "tmp": "tmp"} if args_tss.ta_files is not None: self.tmps["ta"] = os.path.join(args_tss.ta_files, "tmp") else: self.tmps["ta"] = None self.gff_path = os.path.join(args_tss.gffs, "tmp") if args_tss.manual is not None: self.manual_path = os.path.join(args_tss.manual, "tmp") self.wig_path = os.path.join(args_tss.wig_folder, "tmp") self.fasta_path = os.path.join(args_tss.fastas, "tmp") self.stat_outfolder = os.path.join(args_tss.out_folder, "statistics") self.gff_outfolder = os.path.join(args_tss.out_folder, "gffs") def _assign_dict(self, lib_datas): return {"wig": lib_datas[0], "tex": lib_datas[1], "condition": int(lib_datas[2]), "replicate": lib_datas[3], "strand": lib_datas[4]} def _print_lib(self, lib_num, lib_list, out, wig_folder, prefix, rep_set): for num_id in range(1, lib_num+1): cond_list = [] for lib in lib_list: if num_id == lib["condition"]: cond_list.append(lib) cond_sort_list = sorted(cond_list, key=lambda k: k['replicate']) reps = [] for cond in cond_sort_list: out.write("{0}_{1}{2} = {3}\n".format( prefix, cond["condition"], cond["replicate"], os.path.join(wig_folder, cond["wig"]))) reps.append(cond["replicate"]) for rep in sorted(rep_set): if rep not in reps: out.write("{0}_{1}{2} = \n".format( prefix, cond["condition"], rep)) def _start_to_run(self, tsspredator_path, config_file, out_path, prefix, log): print("Running TSSpredator for " + prefix) log.write("Make sure the version of TSSpredator is at least 1.06.\n") out = open(os.path.join(out_path, "log.txt"), "w") err = open(os.path.join(out_path, "err.txt"), "w") log.write(" ".join(["java", "-jar", tsspredator_path, config_file]) + "\n") call(["java", "-jar", tsspredator_path, config_file], stdout=out, stderr=err) out.close() err.close() log.write("Done!\n") log.write("The following files are generated in {0}:\n".format(out_path)) for file_ in os.listdir(out_path): log.write("\t" + file_ + "\n") def _import_lib(self, libs, wig_folder, project_strain_name, out, gff, program, fasta): lib_dict = {"fp": [], "fm": [], "nm": [], "np": []} lib_num = 0 rep_set = set() list_num_id = [] for lib in libs: lib_datas = lib.split(":") if not lib_datas[0].endswith(".wig"): print("Error: Wiggle files are not end with .wig!") sys.exit() for wig in os.listdir(wig_folder): filename = wig.split("_STRAIN_") if (filename[0] == lib_datas[0][:-4]) and ( filename[1][:-4] == project_strain_name): lib_datas[0] = wig if int(lib_datas[2]) > lib_num: lib_num = int(lib_datas[2]) if lib_datas[3] not in rep_set: rep_set.add(lib_datas[3]) if (lib_datas[1] == "tex") and (lib_datas[4] == "+"): lib_dict["fp"].append(self._assign_dict(lib_datas)) elif (lib_datas[1] == "tex") and (lib_datas[4] == "-"): lib_dict["fm"].append(self._assign_dict(lib_datas)) elif (lib_datas[1] == "notex") and (lib_datas[4] == "+"): lib_dict["np"].append(self._assign_dict(lib_datas)) elif (lib_datas[1] == "notex") and (lib_datas[4] == "-"): lib_dict["nm"].append(self._assign_dict(lib_datas)) for num_id in range(1, lib_num+1): out.write("annotation_{0} = {1}\n".format(num_id, gff)) if program.lower() == "tss": self._print_lib(lib_num, lib_dict["fm"], out, wig_folder, "fivePrimeMinus", rep_set) self._print_lib(lib_num, lib_dict["fp"], out, wig_folder, "fivePrimePlus", rep_set) elif program.lower() == "ps": self._print_lib(lib_num, lib_dict["nm"], out, wig_folder, "fivePrimeMinus", rep_set) self._print_lib(lib_num, lib_dict["np"], out, wig_folder, "fivePrimePlus", rep_set) else: print("Error: Wrong program name! Please assing tss " "or processing_site.") sys.exit() for num_id in range(1, lib_num+1): out.write("genome_{0} = {1}\n".format(num_id, fasta)) for num_id in range(1, lib_num+1): list_num_id.append(str(num_id)) return lib_num, num_id, rep_set, lib_dict, list_num_id def _print_repmatch(self, args_tss, out): '''check replicate match''' detect_all = False for rep in args_tss.repmatch: if "all" in rep: detect_all = True match = rep.split("_")[-1] out.write("minNumRepMatches = {0}\n".format(match)) break if not detect_all: nums = {} matchs = {} for match in args_tss.repmatch: lib = match.split("_")[0] rep = match.split("_")[-1] matchs[lib] = rep if rep not in nums.keys(): nums[rep] = 1 else: nums[rep] += 1 for rep, num in nums.items(): if num == max(nums.values()): out.write("minNumRepMatches = {0}\n".format(rep)) max_rep = rep break for lib, rep in matchs.items(): if rep != max_rep: out.write("minNumRepMatches_{0} = {1}\n".format( lib, rep)) def _extract_best_para(self, args_tss, prefix, log): detect = False for best_file in os.listdir(args_tss.auto_load): if best_file == "_".join(["best", prefix + ".csv"]): bh = open(os.path.join(args_tss.auto_load, best_file),"r" ) lines = bh.readlines() bh.close() if len(lines[len(lines)-1].split("\t")) < 8: print("Error: some information in {0} is missing. " "It may be due to that \"optimize_tss_ps\" did " "not finish successfully.".format(best_file)) log.write("Error: some information in {0} is missing. " "It may be due to that \"optimize_tss_ps\" did " "not finish successfully.\n".format(best_file)) sys.exit() else: para_info = lines[len(lines)-1].split("\t")[1].split("_") detect_all = all(elem in para_info for elem in ["he", "rh", "fa", "rf", "bh", "ef", "pf"]) if (not detect_all) or (len(para_info) != 14): print("Error: {0} is complete. Some parameters are " "missing!".format(best_file)) log.write("Error: {0} is complete. Some parameters " "are missing!\n".format(best_file)) sys.exit() else: detect = True height = para_info[para_info.index("he") + 1] height_reduction = para_info[ para_info.index("rh") + 1] factor = para_info[para_info.index("fa") + 1] factor_reduction = para_info[ para_info.index("rf") + 1] base_height = para_info[ para_info.index("bh") + 1] enrichment_factor = para_info[ para_info.index("ef") + 1] processing_factor = para_info[ para_info.index("pf") + 1] if detect: return height, height_reduction, factor, factor_reduction, \ base_height, enrichment_factor, processing_factor else: print("Error: No best_{0}.csv can be found in {1}! ".format( prefix, args_tss.auto_load)) log.write("Error: No best_{0}.csv can be found in {1}\n".format( prefix, args_tss.auto_load)) sys.exit() def _get_input_para(self, args_tss, prefix, log): if args_tss.genome_order is None: height = args_tss.height[0] height_reduction = args_tss.height_reduction[0] factor = args_tss.factor[0] factor_reduction = args_tss.factor_reduction[0] base_height = args_tss.base_height[0] enrichment_factor = args_tss.enrichment_factor[0] processing_factor = args_tss.processing_factor[0] else: if prefix not in args_tss.genome_order: print("Error: the parameters for {0} were not assigned!".format( prefix)) log.write("Error: the parameters for {0} were not assigned!\n".format( prefix)) sys.exit() else: index = args_tss.genome_order.index(prefix) height = args_tss.height[index] height_reduction = args_tss.height_reduction[index] factor = args_tss.factor[index] factor_reduction = args_tss.factor_reduction[index] base_height = args_tss.base_height[index] enrichment_factor = args_tss.enrichment_factor[index] processing_factor = args_tss.processing_factor[index] return height, height_reduction, factor, factor_reduction, \ base_height, enrichment_factor, processing_factor def _gen_config(self, project_strain_name, args_tss, gff, wig_folder, fasta, config_file, log): '''generation of config files''' log.write("Generating config files for TSSpredator.\n") if args_tss.auto_load is not None: height, height_reduction, factor, factor_reduction, \ base_height, enrichment_factor, processing_factor = \ self._extract_best_para(args_tss, project_strain_name, log) else: height, height_reduction, factor, factor_reduction, \ base_height, enrichment_factor, processing_factor = \ self._get_input_para(args_tss, project_strain_name, log) master_folder = "MasterTable_" + project_strain_name out_path = os.path.join(self.master, master_folder) self.helper.check_make_folder(out_path) out = open(config_file, "w") out.write("TSSinClusterSelectionMethod = HIGHEST\n") out.write("allowedCompareShift = 1\n") out.write("allowedRepCompareShift = 1\n") lib_num, num_id, rep_set, lib_dict, list_num_id = \ self._import_lib(args_tss.libs, wig_folder, project_strain_name, out, gff, args_tss.program, fasta) out.write("idList = ") out.write(",".join(list_num_id) + "\n") out.write("maxASutrLength = 100\n") out.write("maxGapLengthInGene = 500\n") out.write("maxNormalTo5primeFactor = {0}\n".format( processing_factor)) out.write("maxTSSinClusterDistance = {0}\n".format( args_tss.cluster + 1)) out.write("maxUTRlength = {0}\n".format(args_tss.utr_length)) out.write("min5primeToNormalFactor = {0}\n".format( enrichment_factor)) out.write("minCliffFactor = {0}\n".format(factor)) out.write("minCliffFactorDiscount = {0}\n".format( factor_reduction)) out.write("minCliffHeight = {0}\n".format(height)) out.write("minCliffHeightDiscount = {0}\n".format( height_reduction)) out.write("minNormalHeight = {0}\n".format(base_height)) self._print_repmatch(args_tss, out) out.write("minPlateauLength = 0\n") out.write("mode = cond\n") out.write("normPercentile = 0.9\n") if args_tss.program.lower() == "tss": self._print_lib(lib_num, lib_dict["nm"], out, wig_folder, "normalMinus", rep_set) self._print_lib(lib_num, lib_dict["np"], out, wig_folder, "normalPlus", rep_set) else: self._print_lib(lib_num, lib_dict["fm"], out, wig_folder, "normalMinus", rep_set) self._print_lib(lib_num, lib_dict["fp"], out, wig_folder, "normalPlus", rep_set) out.write("numReplicates = {0}\n".format(len(rep_set))) out.write("numberOfDatasets = {0}\n".format(lib_num)) out.write("outputDirectory = {0}\n".format(out_path)) for prefix_id in range(len(args_tss.output_prefixs)): out.write("outputPrefix_{0} = {1}\n".format( prefix_id + 1, args_tss.output_prefixs[prefix_id])) out.write("projectName = {0}\n".format(project_strain_name)) out.write("superGraphCompatibility = igb\n") out.write("texNormPercentile = 0.5\n") out.write("writeGraphs = 0\n") out.write("writeNocornacFiles = 0\n") log.write("\t" + config_file + " is generated.\n") out.close() def _convert_gff(self, prefixs, args_tss, log): for prefix in prefixs: out_file = os.path.join(self.gff_outfolder, "_".join([ prefix, args_tss.program]) + ".gff") gff_f = open(out_file, "w") out_path = os.path.join(self.master, "_".join([ "MasterTable", prefix])) if "MasterTable.tsv" not in os.listdir(out_path): print("Error: There is not MasterTable file in {0} ".format( out_path)) print("Please check configuration file.") log.write("not MasterTable file is found in {0}\n".format( out_path)) else: if args_tss.program.lower() == "processing": feature = "processing_site" elif args_tss.program.lower() == "tss": feature = "TSS" self.converter.convert_mastertable2gff( os.path.join(out_path, "MasterTable.tsv"), "ANNOgesic", feature, prefix, out_file) log.write("\t" + out_file + "is generated.\n") gff_f.close() def _merge_manual(self, tsss, args_tss): '''if manual detected TSS is provided, it can merge manual detected TSS and TSSpredator predicted TSS''' self.helper.check_make_folder(os.path.join(os.getcwd(), self.tmps["tss"])) for tss in tsss: for gff in os.listdir(args_tss.gffs): if (gff[:-4] == tss) and (".gff" in gff): break filename = "_".join([tss, args_tss.program]) + ".gff" predict = os.path.join(self.gff_outfolder, filename) manual = os.path.join(self.manual_path, tss + ".gff") fasta = os.path.join(self.fasta_path, tss + ".fa") stat_file = "stat_compare_TSSpredator_manual_{0}.csv".format(tss) if os.path.exists(manual): print("Merging and classiflying manually-detected " "TSSs for {0}".format(tss)) merge_manual_predict_tss( predict, stat_file, os.path.join(self.tmps["tss"], filename), os.path.join(args_tss.gffs, gff), args_tss, manual, fasta) if os.path.exists(stat_file): shutil.move(stat_file, os.path.join( args_tss.out_folder, "statistics", tss, stat_file)) self.helper.move_all_content(self.tmps["tss"], self.gff_outfolder, [".gff"]) shutil.rmtree(self.tmps["tss"]) def _validate(self, tsss, args_tss, log): '''validate TSS with genome annotation''' print("Validating TSSs with genome annotations") log.write("Running validate_gene.py to compare genome " "annotations and TSSs/PSs.\n") for tss in tsss: for gff in os.listdir(args_tss.gffs): if (gff[:-4] == tss) and (".gff" in gff): break stat_file = os.path.join( self.stat_outfolder, tss, "".join(["stat_gene_vali_", tss, ".csv"])) out_cds_file = os.path.join(args_tss.out_folder, "tmp.gff") if args_tss.program.lower() == "tss": compare_file = os.path.join(self.gff_outfolder, "_".join([tss, "TSS.gff"])) elif args_tss.program.lower() == "processing": compare_file = os.path.join(self.gff_outfolder, "_".join([tss, "processing.gff"])) validate_gff(compare_file, os.path.join(args_tss.gffs, gff), stat_file, out_cds_file, args_tss.utr_length, args_tss.program.lower()) log.write("\t" + stat_file + " is generated.\n") shutil.move(out_cds_file, os.path.join(args_tss.gffs, gff)) def _compare_ta(self, tsss, args_tss, log): '''compare TSS with transcript''' detect = False log.write("Running stat_TA_comparison to compare transcripts " "and TSSs/PSs.\n") print("Comparing transcripts and TSSs") self.multiparser.parser_gff(args_tss.ta_files, "transcript") self.multiparser.combine_gff(args_tss.gffs, self.tmps["ta"], None, "transcript") for tss in tsss: stat_out = os.path.join( self.stat_outfolder, tss, "".join([ "stat_compare_TSS_transcript_", tss, ".csv"])) for ta in os.listdir(self.tmps["ta"]): filename = ta.split("_transcript") if (filename[0] == tss) and (filename[1] == ".gff"): detect = True break compare_file = os.path.join(self.gff_outfolder, "_".join([tss, "TSS.gff"])) if detect: stat_ta_tss(os.path.join(self.tmps["ta"], ta), compare_file, stat_out, self.tmps["ta_tss"], self.tmps["tss_ta"], args_tss.fuzzy) self.helper.sort_gff(self.tmps["tss_ta"], compare_file) self.helper.sort_gff(self.tmps["ta_tss"], os.path.join(args_tss.ta_files, ta)) os.remove(self.tmps["tss_ta"]) os.remove(self.tmps["ta_tss"]) detect = False log.write("\t" + stat_out + " is generated.\n") def _stat_tss(self, tsss, feature, log): print("Running statistaics") for tss in tsss: compare_file = os.path.join(self.gff_outfolder, "_".join([tss, feature]) + ".gff") stat_tsspredator( compare_file, feature, os.path.join(self.stat_outfolder, tss, "_".join([ "stat", feature, "class", tss]) + ".csv"), os.path.join(self.stat_outfolder, tss, "_".join([ "stat", feature, "libs", tss]) + ".csv")) self.helper.move_all_content(os.getcwd(), os.path.join( self.stat_outfolder, tss), ["_class", ".png"]) if os.path.exists(os.path.join( self.stat_outfolder, "TSSstatistics.tsv")): shutil.move( os.path.join( self.stat_outfolder, "TSSstatistics.tsv"), os.path.join( self.stat_outfolder, tss, "TSSstatistics.tsv")) plot_venn(compare_file, feature) self.helper.move_all_content(os.getcwd(), os.path.join( self.stat_outfolder, tss), ["_venn", ".png"]) log.write("The following files in {0} are generated:\n".format( (os.path.join(self.stat_outfolder, tss)))) for file_ in os.listdir(os.path.join( self.stat_outfolder, tss)): log.write("\t" + file_ + "\n") def _get_prefixs(self, args_tss): prefixs = [] detect = False for fasta in os.listdir(self.fasta_path): run = False for gff in os.listdir(self.gff_path): if fasta[:-3] == gff[:-4]: prefix = fasta[:-3] for wig in os.listdir(self.wig_path): filename = wig.split("_STRAIN_") if filename[1][:-4] == prefix: detect = True break if detect: prefixs.append(prefix) return prefixs def _merge_wigs(self, wig_folder, prefix, libs): self.helper.check_make_folder(os.path.join(os.getcwd(), self.tmps["tmp"])) for wig_file in os.listdir(wig_folder): for lib in libs: info = lib.split(":") if (info[0][:-4] in wig_file) and (info[-1] == "+") and ( prefix in wig_file) and ( os.path.isfile(os.path.join(wig_folder, wig_file))): Helper().merge_file( os.path.join(wig_folder, wig_file), os.path.join("tmp", "merge_forward.wig")) if (info[0][:-4] in wig_file) and (info[-1] == "-") and ( prefix in wig_file) and ( os.path.isfile(os.path.join(wig_folder, wig_file))): Helper().merge_file( os.path.join(wig_folder, wig_file), os.path.join("tmp", "merge_reverse.wig")) def _check_orphan(self, prefixs, wig_folder, args_tss): '''if genome has no locus tag, it can use for classify the TSS''' for prefix in prefixs: self._merge_wigs(wig_folder, prefix, args_tss.libs) tmp_tss = os.path.join(self.tmps["tmp"], "_".join([ prefix, args_tss.program + ".gff"])) pre_tss = os.path.join(self.gff_outfolder, "_".join([ prefix, args_tss.program + ".gff"])) check_orphan(pre_tss, os.path.join( args_tss.gffs, prefix + ".gff"), "tmp/merge_forward.wig", "tmp/merge_reverse.wig", tmp_tss) shutil.move(tmp_tss, pre_tss) shutil.rmtree("tmp") def _remove_files(self, args_tss): print("Remove temperary files and folders") self.helper.remove_tmp_dir(args_tss.fastas) self.helper.remove_tmp_dir(args_tss.gffs) self.helper.remove_tmp_dir(args_tss.ta_files) if "merge_forward.wig" in os.listdir(os.getcwd()): os.remove("merge_forward.wig") if "merge_reverse.wig" in os.listdir(os.getcwd()): os.remove("merge_reverse.wig") shutil.rmtree(args_tss.wig_folder) if args_tss.manual is not None: shutil.rmtree(args_tss.manual) def _deal_with_overlap(self, out_folder, args_tss): '''deal with the situation that TSS and processing site at the same position''' if not args_tss.overlap_feature: pass else: print("Comparing TSSs and Processing sites") if args_tss.program.lower() == "tss": for tss in os.listdir(out_folder): if tss.endswith("_TSS.gff"): ref = self.helper.get_correct_file( args_tss.overlap_gffs, "_processing.gff", tss.replace("_TSS.gff", ""), None, None) filter_tss_pro(os.path.join(out_folder, tss), ref, args_tss.program, args_tss.cluster) elif args_tss.program.lower() == "processing": for tss in os.listdir(out_folder): if tss.endswith("_processing.gff"): ref = self.helper.get_correct_file( args_tss.overlap_gffs, "_TSS.gff", tss.replace("_processing.gff", ""), None, None) filter_tss_pro(os.path.join(out_folder, tss), ref, args_tss.program, args_tss.cluster) def _low_expression(self, args_tss, gff_folder): '''deal with the low expressed TSS''' prefix = None self._merge_wigs(args_tss.wig_folder, "wig", args_tss.libs) for gff in os.listdir(gff_folder): if (args_tss.program.lower() == "tss") and ( gff.endswith("_TSS.gff")): prefix = gff.replace("_TSS.gff", "") elif (args_tss.program.lower() == "processing") and ( gff.endswith("_processing.gff")): prefix = gff.replace("_processing.gff", "") if prefix: out = open(os.path.join( self.stat_outfolder, prefix, "_".join([ "stat", prefix, "low_expression_cutoff.csv"])), "w") out.write("\t".join(["Genome", "Cutoff_coverage"]) + "\n") cutoff = filter_low_expression( os.path.join(gff_folder, gff), args_tss, "tmp/merge_forward.wig", "tmp/merge_reverse.wig", "tmp/without_low_expression.gff") out.write("\t".join([prefix, str(cutoff)]) + "\n") os.remove(os.path.join(gff_folder, gff)) shutil.move("tmp/without_low_expression.gff", os.path.join(gff_folder, gff)) prefix = None out.close() def run_tsspredator(self, args_tss, log): input_folder = os.path.join(args_tss.out_folder, "configs") for gff in os.listdir(args_tss.gffs): if gff.endswith(".gff"): self.helper.check_uni_attributes(os.path.join( args_tss.gffs, gff)) self.helper.check_make_folder(self.gff_outfolder) self.multiparser.parser_fasta(args_tss.fastas) self.multiparser.parser_gff(args_tss.gffs, None) self.multiparser.parser_wig(args_tss.wig_folder) prefixs = self._get_prefixs(args_tss) for prefix in prefixs: config = os.path.join(input_folder, "_".join(["config", prefix]) + ".ini") self._gen_config( prefix, args_tss, os.path.join(self.gff_path, prefix + ".gff"), self.wig_path, os.path.join(self.fasta_path, prefix + ".fa"), config, log) out_path = os.path.join( self.master, "_".join(["MasterTable", prefix])) config_file = os.path.join( input_folder, "_".join(["config", prefix]) + ".ini") self._start_to_run(args_tss.tsspredator_path, config_file, out_path, prefix, log) if os.path.exists(os.path.join(out_path, "TSSstatistics.tsv")): shutil.move(os.path.join(out_path, "TSSstatistics.tsv"), os.path.join( self.stat_outfolder, "TSSstatistics.tsv")) if args_tss.program.lower() == "ps": args_tss.program = "processing" self._convert_gff(prefixs, args_tss, log) if args_tss.check_orphan: print("checking the orphan TSSs") log.write("Running check_orphan.py to re-check orphan TSSs.\n") self._check_orphan(prefixs, os.path.join(args_tss.wig_folder, "tmp"), args_tss) self.multiparser.combine_gff(args_tss.gffs, self.gff_outfolder, None, args_tss.program) datas = [] for gff in os.listdir(self.gff_outfolder): if gff.endswith(".gff"): gff_folder = gff.replace("".join(["_", args_tss.program, ".gff"]), "") self.helper.check_make_folder( os.path.join(self.stat_outfolder, gff_folder)) datas.append(gff_folder) if args_tss.remove_low_expression is not None: log.write("Running filter_low_expression.py to filter out " "low expressed TSS/PS.\n") self._low_expression(args_tss, self.gff_outfolder) if args_tss.manual is not None: self.multiparser.parser_gff(args_tss.manual, None) self.multiparser.combine_gff(args_tss.gffs, self.manual_path, None, None) self.multiparser.combine_fasta(args_tss.gffs, self.fasta_path, None) self.multiparser.combine_wig(args_tss.gffs, self.wig_path, None, args_tss.libs) log.write("Running merge_manual.py to merge the manual TSSs.\n") self._merge_manual(datas, args_tss) log.write("Running filter_TSS_pro.py to deal with the overlap " "position between TSS and PS.\n") self._deal_with_overlap(self.gff_outfolder, args_tss) log.write("Running stat_TSSpredator.py to do statistics.\n") self._stat_tss(datas, args_tss.program, log) if args_tss.validate: self._validate(datas, args_tss, log) if args_tss.ta_files is not None: self._compare_ta(datas, args_tss, log) self._remove_files(args_tss)
class TestConverter(unittest.TestCase): def setUp(self): self.converter = Converter() self.example = Example() self.converter.gff3parser = Mock_gff3_parser self.converter._print_rntptt_title = Mock_func().print_rntptt_title self.converter.tsspredator = Mock_TSSPredatorReader() self.converter._read_file = Mock_func().mock_read_file self.gff_file = self.example.gff_file self.ptt_out = self.example.ptt_out self.rnt_out = self.example.rnt_out self.srna_out = self.example.srna_out self.embl_file = self.example.embl_file self.embl_out = self.example.embl_out self.multi_embl = self.example.multi_embl self.gff_out = self.example.gff_out self.mastertable = self.example.mastertable self.tss_file = self.example.tss_file self.fasta_file = self.example.fasta_file self.transterm = self.example.transterm self.term_file = self.example.term_file self.circ_file = self.example.circrna_table self.circ_all = self.example.circrna_all self.circ_best = self.example.circrna_best self.test_folder = "test_folder" self.mock_args = MockClass() if (not os.path.exists(self.test_folder)): os.mkdir(self.test_folder) def tearDown(self): if os.path.exists(self.test_folder): shutil.rmtree(self.test_folder) def test_print_rntptt_file(self): cdss = [] genes = [] rnas = [] gff_dict = Example().gff_dict for gff in gff_dict: if gff["feature"] == "gene": genes.append(self.converter.gff3parser.entries(self, gff)) elif gff["feature"] == "CDS": cdss.append(self.converter.gff3parser.entries(self, gff)) elif gff["feature"] == "tRNA": rnas.append(self.converter.gff3parser.entries(self, gff)) out_p = StringIO() out_r = StringIO() self.converter._print_rntptt_file(out_p, cdss, genes) self.converter._print_rntptt_file(out_r, rnas, genes) self.assertEqual(out_p.getvalue().split("\n")[:-1], self.example.ptt_out_list) self.assertEqual(out_r.getvalue().split("\n")[:-1], self.example.rnt_out_list) out_p.close() out_r.close() def test_srna2pttrnt(self): srna_input_file = os.path.join(self.test_folder, "srna.gff") srna_output_file = os.path.join(self.test_folder, "srna.out") with open(srna_input_file, "w") as fh: fh.write(self.gff_file) srnas = [] self.converter._srna2rntptt(srna_input_file, srna_output_file, srnas, 1234567) datas = import_data(srna_output_file) self.assertEqual(set(datas), set(self.srna_out.split("\n"))) def test_multi_embl_pos(self): embls = [] for line in self.embl_file.split("\n"): datas = self.converter._multi_embl_pos(line.strip()) if datas != "Wrong": embls.append(datas) for index in range(0, 7): self.assertDictEqual(embls[index], self.embl_out[index]) for index in range(0, 2): self.assertDictEqual(embls[-1]["pos"][index], self.multi_embl[index]) def test_parser_embl_data(self): embl_file = os.path.join(self.test_folder, "test.embl") embl_out = os.path.join(self.test_folder, "test.embl_out") out = StringIO() with open(embl_file, "w") as eh: for line in self.embl_file.split("\n"): eh.write(line + "\n") info = self.converter._parser_embl_data(embl_file, out) datas = out.getvalue().split("\n") self.assertEqual(set(datas[:-1]), set(self.gff_out.split("\n"))) self.assertEqual(info[0], "NC_007795.1") for index in range(0, 2): self.assertDictEqual(info[1]["pos"][index], self.multi_embl[index]) out.close() def test_multi_tss_class(self): nums = {"tss": 0, "tss_uni": 0, "class": 1} utrs = {"total": [], "pri": [], "sec": []} tss_features = {"tss_types": [], "locus_tags": [], "utr_lengths": []} tss_index = defaultdict(lambda: 0) master_file = os.path.join(self.test_folder, "test.tsv") fh = StringIO(self.mastertable) for tss in self.converter.tsspredator.entries(fh): self.converter._multi_tss_class( tss, tss_index, tss_features, nums, utrs) fh.close() self.assertDictEqual(nums, {'tss_uni': 0, 'class': 5, 'tss': 2}) def test_convert_mastertable2gff(self): master_file = os.path.join(self.test_folder, "test.tsv") with open(master_file, "w") as th: th.write(self.mastertable) out_gff = os.path.join(self.test_folder, "test.tsv_out") self.converter.convert_mastertable2gff(master_file, "ANNOgesic", "TSS", "aaa", out_gff) datas = import_data(out_gff) self.assertEqual(set(datas), set(self.tss_file.split("\n"))) def test_convert_gff2rntptt(self): srna_input_file = os.path.join(self.test_folder, "srna.gff") srna_output_file = os.path.join(self.test_folder, "srna.out") gff_file = os.path.join(self.test_folder, "test.gff") rnt_file = os.path.join(self.test_folder, "test.rnt") ptt_file = os.path.join(self.test_folder, "test.ptt") fasta_file = os.path.join(self.test_folder, "test.fa") with open(srna_input_file, "w") as fh: fh.write(self.gff_file) with open(gff_file, "w") as fh: fh.write(self.gff_file) with open(fasta_file, "w") as fh: fh.write(self.fasta_file) self.converter.convert_gff2rntptt( gff_file, fasta_file, ptt_file, rnt_file, srna_input_file, srna_output_file) self.assertTrue(srna_output_file) self.assertTrue(rnt_file) self.assertTrue(ptt_file) def test_convert_embl2gff(self): embl_file = os.path.join(self.test_folder, "test.embl") gff_file = os.path.join(self.test_folder, "test.embl_out") with open(embl_file, "w") as eh: for line in self.embl_file.split("\n"): eh.write(line + "\n") self.converter.convert_embl2gff(embl_file, gff_file) datas = import_data(gff_file) self.assertEqual(set(datas[1:-2]), set(self.gff_out.split("\n"))) def test_convert_transtermhp2gff(self): transterm_file = os.path.join( self.test_folder, "test_best_terminator_after_gene.bag") gff_file = os.path.join(self.test_folder, "transterm.gff") with open(transterm_file, "w") as th: th.write(self.transterm) self.converter.convert_transtermhp2gff(transterm_file, gff_file) datas = import_data(gff_file) self.assertEqual(set(datas), set(self.term_file.split("\n"))) def get_info(datas): f_datas = [] for data in datas: if not data.startswith("#"): f_datas.append("\t".join(data.split("\t")[:8])) return f_datas def test_convert_circ2gff(self): circ_file = os.path.join(self.test_folder, "circ.csv") out_all = os.path.join(self.test_folder, "all.gff") out_filter = os.path.join(self.test_folder, "best.gff") with open(circ_file, "w") as ch: ch.write(self.circ_file) args = self.mock_args.mock() args.start_ratio = 0.5 args.end_ratio = 0.5 args.support = 5 self.converter.convert_circ2gff(circ_file, args, out_all, out_filter) datas = import_data(out_all) f_datas = [] for data in datas: if not data.startswith("#"): f_datas.append("\t".join(data.split("\t")[:8])) c_datas = [] for data in self.circ_all.split("\n"): if not data.startswith("#"): c_datas.append("\t".join(data.split("\t")[:8])) self.assertListEqual(f_datas, c_datas) datas = import_data(out_filter) f_datas = [] for data in datas: if not data.startswith("#"): f_datas.append("\t".join(data.split("\t")[:8])) c_datas = [] for data in self.circ_best.split("\n"): if not data.startswith("#"): c_datas.append("\t".join(data.split("\t")[:8])) self.assertListEqual(f_datas, c_datas)
class TSSpredator(object): def __init__(self, args_tss): self.multiparser = Multiparser() self.helper = Helper() self.converter = Converter() self.master = os.path.join(args_tss.out_folder, "MasterTables") self.tmps = {"tss": "tmp_TSS", "ta_tss": "tmp_ta_tss", "tss_ta": "tmp_tss", "tmp": "tmp"} if args_tss.ta_files is not None: self.tmps["ta"] = os.path.join(args_tss.ta_files, "tmp") else: self.tmps["ta"] = None self.gff_path = os.path.join(args_tss.gffs, "tmp") if args_tss.manual is not None: self.manual_path = os.path.join(args_tss.manual, "tmp") self.wig_path = os.path.join(args_tss.wig_folder, "tmp") self.fasta_path = os.path.join(args_tss.fastas, "tmp") self.stat_outfolder = os.path.join(args_tss.out_folder, "statistics") self.gff_outfolder = os.path.join(args_tss.out_folder, "gffs") def _assign_dict(self, lib_datas): return {"wig": lib_datas[0], "tex": lib_datas[1], "condition": int(lib_datas[2]), "replicate": lib_datas[3], "strand": lib_datas[4]} def _print_lib(self, lib_num, lib_list, out, wig_folder, prefix, rep_set): for num_id in range(1, lib_num+1): cond_list = [] for lib in lib_list: if num_id == lib["condition"]: cond_list.append(lib) cond_sort_list = sorted(cond_list, key=lambda k: k['replicate']) reps = [] for cond in cond_sort_list: out.write("{0}_{1}{2} = {3}\n".format( prefix, cond["condition"], cond["replicate"], os.path.join(wig_folder, cond["wig"]))) reps.append(cond["replicate"]) for rep in sorted(rep_set): if rep not in reps: out.write("{0}_{1}{2} = \n".format( prefix, cond["condition"], rep)) def _start_to_run(self, tsspredator_path, config_file, out_path, prefix, log): print("Running TSSpredator for " + prefix) log.write("Make sure the version of TSSpredator is at least 1.06.\n") out = open(os.path.join(out_path, "log.txt"), "w") err = open(os.path.join(out_path, "err.txt"), "w") log.write(" ".join(["java", "-jar", tsspredator_path, config_file]) + "\n") call(["java", "-jar", tsspredator_path, config_file], stdout=out, stderr=err) out.close() err.close() log.write("Done!\n") log.write("The following files are generated in {0}:\n".format(out_path)) for file_ in os.listdir(out_path): log.write("\t" + file_ + "\n") def _import_lib(self, libs, wig_folder, project_strain_name, out, gff, program, fasta): lib_dict = {"fp": [], "fm": [], "nm": [], "np": []} lib_num = 0 rep_set = set() list_num_id = [] for lib in libs: lib_datas = lib.split(":") if not lib_datas[0].endswith(".wig"): print("Error: Wiggle files are not end with .wig!") sys.exit() for wig in os.listdir(wig_folder): filename = wig.split("_STRAIN_") if (filename[0] == lib_datas[0][:-4]) and ( filename[1][:-4] == project_strain_name): lib_datas[0] = wig if int(lib_datas[2]) > lib_num: lib_num = int(lib_datas[2]) if lib_datas[3] not in rep_set: rep_set.add(lib_datas[3]) if (lib_datas[1] == "tex") and (lib_datas[4] == "+"): lib_dict["fp"].append(self._assign_dict(lib_datas)) elif (lib_datas[1] == "tex") and (lib_datas[4] == "-"): lib_dict["fm"].append(self._assign_dict(lib_datas)) elif (lib_datas[1] == "notex") and (lib_datas[4] == "+"): lib_dict["np"].append(self._assign_dict(lib_datas)) elif (lib_datas[1] == "notex") and (lib_datas[4] == "-"): lib_dict["nm"].append(self._assign_dict(lib_datas)) for num_id in range(1, lib_num+1): out.write("annotation_{0} = {1}\n".format(num_id, gff)) if program.lower() == "tss": self._print_lib(lib_num, lib_dict["fm"], out, wig_folder, "fivePrimeMinus", rep_set) self._print_lib(lib_num, lib_dict["fp"], out, wig_folder, "fivePrimePlus", rep_set) elif program.lower() == "ps": self._print_lib(lib_num, lib_dict["nm"], out, wig_folder, "fivePrimeMinus", rep_set) self._print_lib(lib_num, lib_dict["np"], out, wig_folder, "fivePrimePlus", rep_set) else: print("Error: Wrong program name! Please assing tss " "or processing_site.") sys.exit() for num_id in range(1, lib_num+1): out.write("genome_{0} = {1}\n".format(num_id, fasta)) for num_id in range(1, lib_num+1): list_num_id.append(str(num_id)) return lib_num, num_id, rep_set, lib_dict, list_num_id def _print_repmatch(self, args_tss, out): '''check replicate match''' detect_all = False for rep in args_tss.repmatch: if "all" in rep: detect_all = True match = rep.split("_")[-1] out.write("minNumRepMatches = {0}\n".format(match)) break if not detect_all: nums = {} matchs = {} for match in args_tss.repmatch: lib = match.split("_")[0] rep = match.split("_")[-1] matchs[lib] = rep if rep not in nums.keys(): nums[rep] = 1 else: nums[rep] += 1 for rep, num in nums.items(): if num == max(nums.values()): out.write("minNumRepMatches = {0}\n".format(rep)) max_rep = rep break for lib, rep in matchs.items(): if rep != max_rep: out.write("minNumRepMatches_{0} = {1}\n".format( lib, rep)) def _gen_config(self, project_strain_name, args_tss, gff, wig_folder, fasta, config_file, log): '''generation of config files''' master_folder = "MasterTable_" + project_strain_name out_path = os.path.join(self.master, master_folder) self.helper.check_make_folder(out_path) out = open(config_file, "w") out.write("TSSinClusterSelectionMethod = HIGHEST\n") out.write("allowedCompareShift = 1\n") out.write("allowedRepCompareShift = 1\n") lib_num, num_id, rep_set, lib_dict, list_num_id = \ self._import_lib(args_tss.libs, wig_folder, project_strain_name, out, gff, args_tss.program, fasta) out.write("idList = ") out.write(",".join(list_num_id) + "\n") out.write("maxASutrLength = 100\n") out.write("maxGapLengthInGene = 500\n") out.write("maxNormalTo5primeFactor = {0}\n".format( args_tss.processing_factor)) out.write("maxTSSinClusterDistance = {0}\n".format( args_tss.cluster + 1)) out.write("maxUTRlength = {0}\n".format(args_tss.utr_length)) out.write("min5primeToNormalFactor = {0}\n".format( args_tss.enrichment_factor)) out.write("minCliffFactor = {0}\n".format(args_tss.factor)) out.write("minCliffFactorDiscount = {0}\n".format( args_tss.factor_reduction)) out.write("minCliffHeight = {0}\n".format(args_tss.height)) out.write("minCliffHeightDiscount = {0}\n".format( args_tss.height_reduction)) out.write("minNormalHeight = {0}\n".format(args_tss.base_height)) self._print_repmatch(args_tss, out) out.write("minPlateauLength = 0\n") out.write("mode = cond\n") out.write("normPercentile = 0.9\n") if args_tss.program.lower() == "tss": self._print_lib(lib_num, lib_dict["nm"], out, wig_folder, "normalMinus", rep_set) self._print_lib(lib_num, lib_dict["np"], out, wig_folder, "normalPlus", rep_set) else: self._print_lib(lib_num, lib_dict["fm"], out, wig_folder, "normalMinus", rep_set) self._print_lib(lib_num, lib_dict["fp"], out, wig_folder, "normalPlus", rep_set) out.write("numReplicates = {0}\n".format(len(rep_set))) out.write("numberOfDatasets = {0}\n".format(lib_num)) out.write("outputDirectory = {0}\n".format(out_path)) for prefix_id in range(len(args_tss.output_prefixs)): out.write("outputPrefix_{0} = {1}\n".format( prefix_id + 1, args_tss.output_prefixs[prefix_id])) out.write("projectName = {0}\n".format(project_strain_name)) out.write("superGraphCompatibility = igb\n") out.write("texNormPercentile = 0.5\n") out.write("writeGraphs = 0\n") out.write("writeNocornacFiles = 0\n") log.write("\t" + config_file + " is generated.\n") out.close() def _convert_gff(self, prefixs, args_tss, log): for prefix in prefixs: out_file = os.path.join(self.gff_outfolder, "_".join([ prefix, args_tss.program]) + ".gff") gff_f = open(out_file, "w") out_path = os.path.join(self.master, "_".join([ "MasterTable", prefix])) if "MasterTable.tsv" not in os.listdir(out_path): print("Error: There is not MasterTable file in {0} ".format( out_path)) print("Please check configuration file.") log.write("not MasterTable file is found in {0}\n".format( out_path)) else: if args_tss.program.lower() == "processing": feature = "processing_site" elif args_tss.program.lower() == "tss": feature = "TSS" self.converter.convert_mastertable2gff( os.path.join(out_path, "MasterTable.tsv"), "ANNOgesic", feature, prefix, out_file) log.write("\t" + out_file + "is generated.\n") gff_f.close() def _merge_manual(self, tsss, args_tss): '''if manual detected TSS is provided, it can merge manual detected TSS and TSSpredator predicted TSS''' self.helper.check_make_folder(os.path.join(os.getcwd(), self.tmps["tss"])) for tss in tsss: for gff in os.listdir(args_tss.gffs): if (gff[:-4] == tss) and (".gff" in gff): break filename = "_".join([tss, args_tss.program]) + ".gff" predict = os.path.join(self.gff_outfolder, filename) manual = os.path.join(self.manual_path, tss + ".gff") fasta = os.path.join(self.fasta_path, tss + ".fa") stat_file = "stat_compare_TSSpredator_manual_{0}.csv".format(tss) if os.path.exists(manual): print("Merging and classiflying manually-detected " "TSSs for {0}".format(tss)) merge_manual_predict_tss( predict, stat_file, os.path.join(self.tmps["tss"], filename), os.path.join(args_tss.gffs, gff), args_tss, manual, fasta) if os.path.exists(stat_file): shutil.move(stat_file, os.path.join( args_tss.out_folder, "statistics", tss, stat_file)) self.helper.move_all_content(self.tmps["tss"], self.gff_outfolder, [".gff"]) shutil.rmtree(self.tmps["tss"]) def _validate(self, tsss, args_tss, log): '''validate TSS with genome annotation''' print("Validating TSSs with genome annotations") log.write("Running validate_gene.py to compare genome " "annotations and TSSs/PSs.\n") for tss in tsss: for gff in os.listdir(args_tss.gffs): if (gff[:-4] == tss) and (".gff" in gff): break stat_file = os.path.join( self.stat_outfolder, tss, "".join(["stat_gene_vali_", tss, ".csv"])) out_cds_file = os.path.join(args_tss.out_folder, "tmp.gff") if args_tss.program.lower() == "tss": compare_file = os.path.join(self.gff_outfolder, "_".join([tss, "TSS.gff"])) elif args_tss.program.lower() == "processing": compare_file = os.path.join(self.gff_outfolder, "_".join([tss, "processing.gff"])) validate_gff(compare_file, os.path.join(args_tss.gffs, gff), stat_file, out_cds_file, args_tss.utr_length, args_tss.program.lower()) log.write("\t" + stat_file + " is generated.\n") shutil.move(out_cds_file, os.path.join(args_tss.gffs, gff)) def _compare_ta(self, tsss, args_tss, log): '''compare TSS with transcript''' detect = False log.write("Running stat_TA_comparison to compare transcripts " "and TSSs/PSs.\n") print("Comparing transcripts and TSSs") self.multiparser.parser_gff(args_tss.ta_files, "transcript") self.multiparser.combine_gff(args_tss.gffs, self.tmps["ta"], None, "transcript") for tss in tsss: stat_out = os.path.join( self.stat_outfolder, tss, "".join([ "stat_compare_TSS_transcript_", tss, ".csv"])) for ta in os.listdir(self.tmps["ta"]): filename = ta.split("_transcript") if (filename[0] == tss) and (filename[1] == ".gff"): detect = True break compare_file = os.path.join(self.gff_outfolder, "_".join([tss, "TSS.gff"])) if detect: stat_ta_tss(os.path.join(self.tmps["ta"], ta), compare_file, stat_out, self.tmps["ta_tss"], self.tmps["tss_ta"], args_tss.fuzzy) self.helper.sort_gff(self.tmps["tss_ta"], compare_file) self.helper.sort_gff(self.tmps["ta_tss"], os.path.join(args_tss.ta_files, ta)) os.remove(self.tmps["tss_ta"]) os.remove(self.tmps["ta_tss"]) detect = False log.write("\t" + stat_out + " is generated.\n") def _stat_tss(self, tsss, feature, log): print("Running statistaics") for tss in tsss: compare_file = os.path.join(self.gff_outfolder, "_".join([tss, feature]) + ".gff") stat_tsspredator( compare_file, feature, os.path.join(self.stat_outfolder, tss, "_".join([ "stat", feature, "class", tss]) + ".csv"), os.path.join(self.stat_outfolder, tss, "_".join([ "stat", feature, "libs", tss]) + ".csv")) self.helper.move_all_content(os.getcwd(), os.path.join( self.stat_outfolder, tss), ["_class", ".png"]) if os.path.exists(os.path.join( self.stat_outfolder, "TSSstatistics.tsv")): shutil.move( os.path.join( self.stat_outfolder, "TSSstatistics.tsv"), os.path.join( self.stat_outfolder, tss, "TSSstatistics.tsv")) plot_venn(compare_file, feature) self.helper.move_all_content(os.getcwd(), os.path.join( self.stat_outfolder, tss), ["_venn", ".png"]) log.write("The following files in {0} are generated:\n".format( (os.path.join(self.stat_outfolder, tss)))) for file_ in os.listdir(os.path.join( self.stat_outfolder, tss)): log.write("\t" + file_ + "\n") def _set_gen_config(self, args_tss, input_folder, log): prefixs = [] detect = False log.write("Generating config files for TSSpredator.\n") for fasta in os.listdir(self.fasta_path): run = False for gff in os.listdir(self.gff_path): if fasta[:-3] == gff[:-4]: prefix = fasta[:-3] for wig in os.listdir(self.wig_path): filename = wig.split("_STRAIN_") if filename[1][:-4] == prefix: detect = True break if detect: prefixs.append(prefix) config = os.path.join( input_folder, "_".join(["config", prefix]) + ".ini") self._gen_config( prefix, args_tss, os.path.join(self.gff_path, gff), self.wig_path, os.path.join(self.fasta_path, fasta), config, log) return prefixs def _merge_wigs(self, wig_folder, prefix, libs): self.helper.check_make_folder(os.path.join(os.getcwd(), self.tmps["tmp"])) for wig_file in os.listdir(wig_folder): for lib in libs: info = lib.split(":") if (info[0][:-4] in wig_file) and (info[-1] == "+") and ( prefix in wig_file) and ( os.path.isfile(os.path.join(wig_folder, wig_file))): Helper().merge_file( os.path.join(wig_folder, wig_file), os.path.join("tmp", "merge_forward.wig")) if (info[0][:-4] in wig_file) and (info[-1] == "-") and ( prefix in wig_file) and ( os.path.isfile(os.path.join(wig_folder, wig_file))): Helper().merge_file( os.path.join(wig_folder, wig_file), os.path.join("tmp", "merge_reverse.wig")) def _check_orphan(self, prefixs, wig_folder, args_tss): '''if genome has no locus tag, it can use for classify the TSS''' for prefix in prefixs: self._merge_wigs(wig_folder, prefix, args_tss.libs) tmp_tss = os.path.join(self.tmps["tmp"], "_".join([ prefix, args_tss.program + ".gff"])) pre_tss = os.path.join(self.gff_outfolder, "_".join([ prefix, args_tss.program + ".gff"])) check_orphan(pre_tss, os.path.join( args_tss.gffs, prefix + ".gff"), "tmp/merge_forward.wig", "tmp/merge_reverse.wig", tmp_tss) shutil.move(tmp_tss, pre_tss) shutil.rmtree("tmp") def _remove_files(self, args_tss): print("Remove temperary files and folders") self.helper.remove_tmp_dir(args_tss.fastas) self.helper.remove_tmp_dir(args_tss.gffs) self.helper.remove_tmp_dir(args_tss.ta_files) if "merge_forward.wig" in os.listdir(os.getcwd()): os.remove("merge_forward.wig") if "merge_reverse.wig" in os.listdir(os.getcwd()): os.remove("merge_reverse.wig") shutil.rmtree(args_tss.wig_folder) if args_tss.manual is not None: shutil.rmtree(args_tss.manual) def _deal_with_overlap(self, out_folder, args_tss): '''deal with the situation that TSS and processing site at the same position''' if not args_tss.overlap_feature: pass else: print("Comparing TSSs and Processing sites") if args_tss.program.lower() == "tss": for tss in os.listdir(out_folder): if tss.endswith("_TSS.gff"): ref = self.helper.get_correct_file( args_tss.overlap_gffs, "_processing.gff", tss.replace("_TSS.gff", ""), None, None) filter_tss_pro(os.path.join(out_folder, tss), ref, args_tss.program, args_tss.cluster) elif args_tss.program.lower() == "processing": for tss in os.listdir(out_folder): if tss.endswith("_processing.gff"): ref = self.helper.get_correct_file( args_tss.overlap_gffs, "_TSS.gff", tss.replace("_processing.gff", ""), None, None) filter_tss_pro(os.path.join(out_folder, tss), ref, args_tss.program, args_tss.cluster) def _low_expression(self, args_tss, gff_folder): '''deal with the low expressed TSS''' prefix = None self._merge_wigs(args_tss.wig_folder, "wig", args_tss.libs) for gff in os.listdir(gff_folder): if (args_tss.program.lower() == "tss") and ( gff.endswith("_TSS.gff")): prefix = gff.replace("_TSS.gff", "") elif (args_tss.program.lower() == "processing") and ( gff.endswith("_processing.gff")): prefix = gff.replace("_processing.gff", "") if prefix: out = open(os.path.join( self.stat_outfolder, prefix, "_".join([ "stat", prefix, "low_expression_cutoff.csv"])), "w") out.write("\t".join(["Genome", "Cutoff_coverage"]) + "\n") cutoff = filter_low_expression( os.path.join(gff_folder, gff), args_tss, "tmp/merge_forward.wig", "tmp/merge_reverse.wig", "tmp/without_low_expression.gff") out.write("\t".join([prefix, str(cutoff)]) + "\n") os.remove(os.path.join(gff_folder, gff)) shutil.move("tmp/without_low_expression.gff", os.path.join(gff_folder, gff)) prefix = None out.close() def run_tsspredator(self, args_tss, log): input_folder = os.path.join(args_tss.out_folder, "configs") for gff in os.listdir(args_tss.gffs): if gff.endswith(".gff"): self.helper.check_uni_attributes(os.path.join( args_tss.gffs, gff)) self.helper.check_make_folder(self.gff_outfolder) self.multiparser.parser_fasta(args_tss.fastas) self.multiparser.parser_gff(args_tss.gffs, None) self.multiparser.parser_wig(args_tss.wig_folder) prefixs = self._set_gen_config(args_tss, input_folder, log) for prefix in prefixs: out_path = os.path.join( self.master, "_".join(["MasterTable", prefix])) config_file = os.path.join( input_folder, "_".join(["config", prefix]) + ".ini") self._start_to_run(args_tss.tsspredator_path, config_file, out_path, prefix, log) if os.path.exists(os.path.join(out_path, "TSSstatistics.tsv")): shutil.move(os.path.join(out_path, "TSSstatistics.tsv"), os.path.join( self.stat_outfolder, "TSSstatistics.tsv")) if args_tss.program.lower() == "ps": args_tss.program = "processing" self._convert_gff(prefixs, args_tss, log) if args_tss.check_orphan: print("checking the orphan TSSs") log.write("Running check_orphan.py to re-check orphan TSSs.\n") self._check_orphan(prefixs, os.path.join(args_tss.wig_folder, "tmp"), args_tss) self.multiparser.combine_gff(args_tss.gffs, self.gff_outfolder, None, args_tss.program) datas = [] for gff in os.listdir(self.gff_outfolder): if gff.endswith(".gff"): gff_folder = gff.replace("".join(["_", args_tss.program, ".gff"]), "") self.helper.check_make_folder( os.path.join(self.stat_outfolder, gff_folder)) datas.append(gff_folder) if args_tss.remove_low_expression is not None: log.write("Running filter_low_expression.py to filter out " "low expressed TSS/PS.\n") self._low_expression(args_tss, self.gff_outfolder) if args_tss.manual is not None: self.multiparser.parser_gff(args_tss.manual, None) self.multiparser.combine_gff(args_tss.gffs, self.manual_path, None, None) self.multiparser.combine_fasta(args_tss.gffs, self.fasta_path, None) self.multiparser.combine_wig(args_tss.gffs, self.wig_path, None, args_tss.libs) log.write("Running merge_manual.py to merge the manual TSSs.\n") self._merge_manual(datas, args_tss) log.write("Running filter_TSS_pro.py to deal with the overlap " "position between TSS and PS.\n") self._deal_with_overlap(self.gff_outfolder, args_tss) log.write("Running stat_TSSpredator.py to do statistics.\n") self._stat_tss(datas, args_tss.program, log) if args_tss.validate: self._validate(datas, args_tss, log) if args_tss.ta_files is not None: self._compare_ta(datas, args_tss, log) self._remove_files(args_tss)
class TSSpredator(object): def __init__(self, args_tss): self.multiparser = Multiparser() self.helper = Helper() self.converter = Converter() self.master = os.path.join(args_tss.out_folder, "MasterTables") self.tmps = {"tss": "tmp_TSS", "ta_tss": "tmp_ta_tss", "tss_ta": "tmp_tss", "tmp": "tmp"} if args_tss.ta_files is not None: self.tmps["ta"] = os.path.join(args_tss.ta_files, "tmp") else: self.tmps["ta"] = None self.gff_path = os.path.join(args_tss.gffs, "tmp") self.wig_path = os.path.join(args_tss.wig_folder, "tmp") self.fasta_path = os.path.join(args_tss.fastas, "tmp") self.stat_outfolder = os.path.join(args_tss.out_folder, "statistics") self.gff_outfolder = os.path.join(args_tss.out_folder, "gffs") def _assign_dict(self, lib_datas): return {"wig": lib_datas[0], "tex": lib_datas[1], "condition": int(lib_datas[2]), "replicate": lib_datas[3], "strand": lib_datas[4]} def _print_lib(self, lib_num, lib_list, out, wig_folder, prefix): for num_id in range(1, lib_num+1): cond_list = [] for lib in lib_list: if num_id == lib["condition"]: cond_list.append(lib) cond_sort_list = sorted(cond_list, key=lambda k: k['replicate']) for cond in cond_sort_list: out.write("{0}_{1}{2} = {3}\n".format( prefix, cond["condition"], cond["replicate"], os.path.join(wig_folder, cond["wig"]))) def _start_to_run(self, tsspredator_path, config_file, out_path, prefix): print("Running TSSpredator for " + prefix) out = open(os.path.join(out_path, "log.txt"), "w") err = open(os.path.join(out_path, "err.txt"), "w") call(["java", "-jar", tsspredator_path, config_file], stdout=out, stderr=err) out.close() err.close() def _import_lib(self, libs, wig_folder, project_strain_name, out, gff, program, fasta): lib_dict = {"fp": [], "fm": [], "nm": [], "np": []} lib_num = 0 rep_set = set() list_num_id = [] print("Runniun {0} now...".format(program)) for lib in libs: lib_datas = lib.split(":") if not lib_datas[0].endswith(".wig"): print("Error:Exist a not proper wig files!!") sys.exit() for wig in os.listdir(wig_folder): filename = wig.split("_STRAIN_") if (filename[0] == lib_datas[0][:-4]) and ( filename[1][:-4] == project_strain_name): lib_datas[0] = wig if int(lib_datas[2]) > lib_num: lib_num = int(lib_datas[2]) if lib_datas[3] not in rep_set: rep_set.add(lib_datas[3]) if (lib_datas[1] == "tex") and (lib_datas[4] == "+"): lib_dict["fp"].append(self._assign_dict(lib_datas)) elif (lib_datas[1] == "tex") and (lib_datas[4] == "-"): lib_dict["fm"].append(self._assign_dict(lib_datas)) elif (lib_datas[1] == "notex") and (lib_datas[4] == "+"): lib_dict["np"].append(self._assign_dict(lib_datas)) elif (lib_datas[1] == "notex") and (lib_datas[4] == "-"): lib_dict["nm"].append(self._assign_dict(lib_datas)) for num_id in range(1, lib_num+1): out.write("annotation_{0} = {1}\n".format(num_id, gff)) if program.lower() == "tss": self._print_lib(lib_num, lib_dict["fm"], out, wig_folder, "fivePrimeMinus") self._print_lib(lib_num, lib_dict["fp"], out, wig_folder, "fivePrimePlus") elif program.lower() == "processing_site": self._print_lib(lib_num, lib_dict["nm"], out, wig_folder, "fivePrimeMinus") self._print_lib(lib_num, lib_dict["np"], out, wig_folder, "fivePrimePlus") else: print("Error: Wrong program name!!!") sys.exit() for num_id in range(1, lib_num+1): out.write("genome_{0} = {1}\n".format(num_id, fasta)) for num_id in range(1, lib_num+1): list_num_id.append(str(num_id)) return lib_num, num_id, rep_set, lib_dict, list_num_id def _gen_config(self, project_strain_name, args_tss, gff, wig_folder, fasta, config_file): master_folder = "MasterTable_" + project_strain_name out_path = os.path.join(self.master, master_folder) self.helper.check_make_folder(out_path) out = open(config_file, "w") out.write("TSSinClusterSelectionMethod = HIGHEST\n") out.write("allowedCompareShift = 1\n") out.write("allowedRepCompareShift = 1\n") lib_num, num_id, rep_set, lib_dict, list_num_id = \ self._import_lib(args_tss.libs, wig_folder, project_strain_name, out, gff, args_tss.program, fasta) out.write("idList = ") out.write(",".join(list_num_id) + "\n") out.write("maxASutrLength = 100\n") out.write("maxGapLengthInGene = 500\n") out.write("maxNormalTo5primeFactor = {0}\n".format( args_tss.processing_factor)) out.write("maxTSSinClusterDistance = {0}\n".format( args_tss.cluster + 1)) out.write("maxUTRlength = {0}\n".format(args_tss.utr_length)) out.write("min5primeToNormalFactor = {0}\n".format( args_tss.enrichment_factor)) out.write("minCliffFactor = {0}\n".format(args_tss.factor)) out.write("minCliffFactorDiscount = {0}\n".format( args_tss.factor_reduction)) out.write("minCliffHeight = {0}\n".format(args_tss.height)) out.write("minCliffHeightDiscount = {0}\n".format( args_tss.height_reduction)) out.write("minNormalHeight = {0}\n".format(args_tss.base_height)) out.write("minNumRepMatches = {0}\n".format(args_tss.repmatch)) out.write("minPlateauLength = 0\n") out.write("mode = cond\n") out.write("normPercentile = 0.9\n") if args_tss.program.lower() == "tss": self._print_lib(lib_num, lib_dict["nm"], out, wig_folder, "normalMinus") self._print_lib(lib_num, lib_dict["np"], out, wig_folder, "normalPlus") else: self._print_lib(lib_num, lib_dict["fm"], out, wig_folder, "normalMinus") self._print_lib(lib_num, lib_dict["fp"], out, wig_folder, "normalPlus") out.write("numReplicates = {0}\n".format(len(rep_set))) out.write("numberOfDatasets = {0}\n".format(lib_num)) out.write("outputDirectory = {0}\n".format(out_path)) for prefix_id in range(len(args_tss.output_prefixs)): out.write("outputPrefix_{0} = {1}\n".format( prefix_id + 1, args_tss.output_prefixs[prefix_id])) out.write("projectName = {0}\n".format(project_strain_name)) out.write("superGraphCompatibility = igb\n") out.write("texNormPercentile = 0.5\n") out.write("writeGraphs = 0\n") out.write("writeNocornacFiles = 0\n") out.close() def _convert_gff(self, prefixs, args_tss): for prefix in prefixs: out_file = os.path.join(self.gff_outfolder, "_".join([ prefix, args_tss.program]) + ".gff") gff_f = open(out_file, "w") out_path = os.path.join(self.master, "_".join([ "MasterTable", prefix])) if "MasterTable.tsv" not in os.listdir(out_path): print("Error:there is not MasterTable file in {0}".format( out_path)) print("Please check configuration file.") else: self.converter.convert_mastertable2gff( os.path.join(out_path, "MasterTable.tsv"), "ANNOgesic", args_tss.program, prefix, out_file) gff_f.close() def _merge_manual(self, tsss, args_tss): self.helper.check_make_folder(os.path.join(os.getcwd(), self.tmps["tss"])) for tss in tsss: for gff in os.listdir(args_tss.gffs): if (gff[:-4] == tss) and (".gff" in gff): break filename = "_".join([tss, args_tss.program]) + ".gff" predict = os.path.join(self.gff_outfolder, filename) print("Running merge and classify manual ....") stat_file = "stat_compare_TSSpredator_manual_{0}.csv".format(tss) merge_manual_predict_tss( predict, stat_file, os.path.join(self.tmps["tss"], filename), os.path.join(args_tss.gffs, gff), args_tss) shutil.move(stat_file, os.path.join(args_tss.out_folder, "statistics", tss, stat_file)) self.helper.move_all_content(self.tmps["tss"], self.gff_outfolder, [".gff"]) shutil.rmtree(self.tmps["tss"]) def _validate(self, tsss, args_tss): print("Running validation of annotation....") for tss in tsss: for gff in os.listdir(args_tss.gffs): if (gff[:-4] == tss) and (".gff" in gff): break stat_file = os.path.join( self.stat_outfolder, tss, "".join(["stat_gene_vali_", tss, ".csv"])) out_cds_file = os.path.join(args_tss.out_folder, "tmp.gff") if args_tss.program.lower() == "tss": compare_file = os.path.join(self.gff_outfolder, "_".join([tss, "TSS.gff"])) elif args_tss.program.lower() == "processing": compare_file = os.path.join(self.gff_outfolder, "_".join([tss, "processing.gff"])) validate_gff(compare_file, os.path.join(args_tss.gffs, gff), stat_file, out_cds_file, args_tss.utr_length, args_tss.program.lower()) shutil.move(out_cds_file, os.path.join(args_tss.gffs, gff)) def _compare_ta(self, tsss, args_tss): detect = False print("Running compare transcript assembly and TSS ...") self.multiparser.parser_gff(args_tss.ta_files, "transcript") self.multiparser.combine_gff(args_tss.gffs, self.tmps["ta"], None, "transcript") for tss in tsss: stat_out = os.path.join( self.stat_outfolder, tss, "".join([ "stat_compare_TSS_Transcriptome_assembly_", tss, ".csv"])) for ta in os.listdir(self.tmps["ta"]): filename = ta.split("_transcript") if (filename[0] == tss) and (filename[1] == ".gff"): detect = True break compare_file = os.path.join(self.gff_outfolder, "_".join([tss, "TSS.gff"])) if detect: stat_ta_tss(os.path.join(self.tmps["ta"], ta), compare_file, stat_out, self.tmps["ta_tss"], self.tmps["tss_ta"], args_tss.fuzzy) self.helper.sort_gff(self.tmps["tss_ta"], compare_file) self.helper.sort_gff(self.tmps["ta_tss"], os.path.join(args_tss.ta_files, ta)) os.remove(self.tmps["tss_ta"]) os.remove(self.tmps["ta_tss"]) detect = False def _stat_tss(self, tsss, feature): print("Running statistaics.....") for tss in tsss: compare_file = os.path.join(self.gff_outfolder, "_".join([tss, feature]) + ".gff") stat_tsspredator( compare_file, feature, os.path.join(self.stat_outfolder, tss, "_".join([ "stat", feature, "class", tss]) + ".csv"), os.path.join(self.stat_outfolder, tss, "_".join([ "stat", feature, "libs", tss]) + ".csv")) self.helper.move_all_content(os.getcwd(), os.path.join( self.stat_outfolder, tss), ["_class", ".png"]) if os.path.exists(os.path.join( self.stat_outfolder, "TSSstatistics.tsv")): shutil.move( os.path.join( self.stat_outfolder, "TSSstatistics.tsv"), os.path.join( self.stat_outfolder, tss, "TSSstatistics.tsv")) plot_venn(compare_file, feature) self.helper.move_all_content(os.getcwd(), os.path.join( self.stat_outfolder, tss), ["_venn", ".png"]) def _set_gen_config(self, args_tss, input_folder): prefixs = [] detect = False for fasta in os.listdir(self.fasta_path): for gff in os.listdir(self.gff_path): if fasta[:-3] == gff[:-4]: prefix = fasta[:-3] for wig in os.listdir(self.wig_path): filename = wig.split("_STRAIN_") if filename[1][:-4] == prefix: detect = True break if detect: prefixs.append(prefix) config = os.path.join( input_folder, "_".join(["config", prefix]) + ".ini") self._gen_config( prefix, args_tss, os.path.join(self.gff_path, gff), self.wig_path, os.path.join(self.fasta_path, fasta), config) return prefixs def _merge_wigs(self, wig_folder, prefix, libs): self.helper.check_make_folder(os.path.join(os.getcwd(), self.tmps["tmp"])) for wig_file in os.listdir(wig_folder): for lib in libs: info = lib.split(":") if (info[0][:-4] in wig_file) and (info[-1] == "+") and ( prefix in wig_file) and ( os.path.isfile(os.path.join(wig_folder, wig_file))): Helper().merge_file( os.path.join(wig_folder, wig_file), os.path.join("tmp", "merge_forward.wig")) if (info[0][:-4] in wig_file) and (info[-1] == "-") and ( prefix in wig_file) and ( os.path.isfile(os.path.join(wig_folder, wig_file))): Helper().merge_file( os.path.join(wig_folder, wig_file), os.path.join("tmp", "merge_reverse.wig")) def _check_orphan(self, prefixs, wig_folder, args_tss): for prefix in prefixs: self._merge_wigs(wig_folder, prefix, args_tss.libs) tmp_tss = os.path.join(self.tmps["tmp"], "_".join([ prefix, args_tss.program + ".gff"])) pre_tss = os.path.join(self.gff_outfolder, "_".join([ prefix, args_tss.program + ".gff"])) check_orphan(pre_tss, os.path.join( args_tss.gffs, prefix + ".gff"), "tmp/merge_forward.wig", "tmp/merge_reverse.wig", tmp_tss) shutil.move(tmp_tss, pre_tss) shutil.rmtree("tmp") def _remove_files(self, args_tss): print("Remove temperary files and folders...") self.helper.remove_tmp(args_tss.fastas) self.helper.remove_tmp(args_tss.gffs) self.helper.remove_tmp(args_tss.wig_folder) self.helper.remove_tmp(args_tss.ta_files) if "merge_forward.wig" in os.listdir(os.getcwd()): os.remove("merge_forward.wig") if "merge_reverse.wig" in os.listdir(os.getcwd()): os.remove("merge_reverse.wig") def _deal_with_overlap(self, out_folder, args_tss): if args_tss.overlap_feature.lower() == "both": pass else: print("Comparing TSS and Processing site...") if args_tss.program.lower() == "tss": for tss in os.listdir(out_folder): if tss.endswith("_TSS.gff"): ref = self.helper.get_correct_file( args_tss.references, "_processing.gff", tss.replace("_TSS.gff", ""), None, None) filter_tss_pro(os.path.join(out_folder, tss), ref, args_tss.overlap_feature, args_tss.cluster) elif args_tss.program.lower() == "processing_site": for tss in os.listdir(out_folder): if tss.endswith("_processing.gff"): ref = self.helper.get_correct_file( args_tss.references, "_TSS.gff", tss.replace("_processing.gff", ""), None, None) filter_tss_pro(os.path.join(out_folder, tss), ref, args_tss.overlap_feature, args_tss.cluster) def _low_expression(self, args_tss, gff_folder): prefix = None self._merge_wigs(args_tss.wig_folder, "wig", args_tss.libs) for gff in os.listdir(gff_folder): if (args_tss.program.lower() == "tss") and ( gff.endswith("_TSS.gff")): prefix = gff.replace("_TSS.gff", "") elif (args_tss.program.lower() == "processing") and ( gff.endswith("_processing.gff")): prefix = gff.replace("_processing.gff", "") if prefix: out = open(os.path.join( self.stat_outfolder, prefix, "_".join([ "stat", prefix, "low_expression_cutoff.csv"])), "w") out.write("\t".join(["strain", "cutoff_coverage"]) + "\n") cutoff = filter_low_expression( os.path.join(gff_folder, gff), args_tss, "tmp/merge_forward.wig", "tmp/merge_reverse.wig", "tmp/without_low_expression.gff") out.write("\t".join([prefix, str(cutoff)]) + "\n") os.remove(os.path.join(gff_folder, gff)) shutil.move("tmp/without_low_expression.gff", os.path.join(gff_folder, gff)) prefix = None out.close() def run_tsspredator(self, args_tss): input_folder = os.path.join(args_tss.out_folder, "configs") for gff in os.listdir(args_tss.gffs): if gff.endswith(".gff"): self.helper.check_uni_attributes(os.path.join( args_tss.gffs, gff)) self.helper.check_make_folder(self.gff_outfolder) self.multiparser.parser_fasta(args_tss.fastas) self.multiparser.parser_gff(args_tss.gffs, None) self.multiparser.parser_wig(args_tss.wig_folder) prefixs = self._set_gen_config(args_tss, input_folder) for prefix in prefixs: out_path = os.path.join( self.master, "_".join(["MasterTable", prefix])) config_file = os.path.join( input_folder, "_".join(["config", prefix]) + ".ini") self._start_to_run(args_tss.tsspredator_path, config_file, out_path, prefix) if os.path.exists(os.path.join(out_path, "TSSstatistics.tsv")): shutil.move(os.path.join(out_path, "TSSstatistics.tsv"), os.path.join( self.stat_outfolder, "TSSstatistics.tsv")) if args_tss.program.lower() == "processing_site": args_tss.program = "processing" self._convert_gff(prefixs, args_tss) if args_tss.check_orphan: print("checking the orphan TSS...") self._check_orphan(prefixs, os.path.join(args_tss.wig_folder, "tmp"), args_tss) self.multiparser.combine_gff(args_tss.gffs, self.gff_outfolder, None, args_tss.program) datas = [] for gff in os.listdir(self.gff_outfolder): if gff.endswith(".gff"): gff_folder = gff.replace("".join(["_", args_tss.program, ".gff"]), "") self.helper.check_make_folder( os.path.join(self.stat_outfolder, gff_folder)) datas.append(gff_folder) if args_tss.remove_low_expression is not None: self._low_expression(args_tss, self.gff_outfolder) if args_tss.manual is not None: self.multiparser.combine_wig(args_tss.gffs, self.wig_path, None, args_tss.libs) self._merge_manual(datas, args_tss) self._deal_with_overlap(self.gff_outfolder, args_tss) if args_tss.stat: self._stat_tss(datas, args_tss.program) if args_tss.validate: self._validate(datas, args_tss) if args_tss.ta_files is not None: self._compare_ta(datas, args_tss) self._remove_files(args_tss)