33     cmdline::parser cmd;  in main()  local
35     cmd.add<string>("in1", 'i', "read1 input file name", false, "");  in main()
36     cmd.add<string>("out1", 'o', "read1 output file name", false, "");  in main()
37     cmd.add<string>("in2", 'I', "read2 input file name", false, "");  in main()
38     cmd.add<string>("out2", 'O', "read2 output file name", false, "");  in main()
39 …cmd.add<string>("unpaired1", 0, "for PE input, if read1 passed QC but read2 not, it will be writte…  in main()
40 …cmd.add<string>("unpaired2", 0, "for PE input, if read2 passed QC but read1 not, it will be writte…  in main()
41 …cmd.add<string>("overlapped_out", 0, "for each read pair, output the overlapped region if it has n…  in main()
42 …cmd.add<string>("failed_out", 0, "specify the file to store reads that cannot pass the filters.", …  in main()
43 …cmd.add("merge", 'm', "for paired-end input, merge each pair of reads into a single read if they a…  in main()
44 …cmd.add<string>("merged_out", 0, "in the merging mode, specify the file name to store merged outpu…  in main()
45 …cmd.add("include_unmerged", 0, "in the merging mode, write the unmerged or unpaired reads to the f…  in main()
46 …cmd.add("phred64", '6', "indicate the input is using phred64 scoring (it'll be converted to phred3…  in main()
47 …cmd.add<int>("compression", 'z', "compression level for gzip output (1 ~ 9). 1 is fastest, 9 is sm…  in main()
48 …cmd.add("stdin", 0, "input from STDIN. If the STDIN is interleaved paired-end FASTQ, please also a…  in main()
49 …cmd.add("stdout", 0, "stream passing-filters reads to STDOUT. This option will result in interleav…  in main()
50 …cmd.add("interleaved_in", 0, "indicate that <in1> is an interleaved FASTQ which contains both read…  in main()
51 …cmd.add<int>("reads_to_process", 0, "specify how many reads/pairs to be processed. Default 0 means…  in main()
52 …cmd.add("dont_overwrite", 0, "don't overwrite existing files. Overwritting is allowed by default."…  in main()
53 …cmd.add("fix_mgi_id", 0, "the MGI FASTQ ID format is not compatible with many BAM operation tools,…  in main()
54 …cmd.add("verbose", 'V', "output verbose log information (i.e. when every 1M reads are processed)."…  in main()
57 …cmd.add("disable_adapter_trimming", 'A', "adapter trimming is enabled by default. If this option i…  in main()
58 …cmd.add<string>("adapter_sequence", 'a', "the adapter for read1. For SE data, if not specified, th…  in main()
59 …cmd.add<string>("adapter_sequence_r2", 0, "the adapter for read2 (PE data only). This is used if R…  in main()
60 …cmd.add<string>("adapter_fasta", 0, "specify a FASTA file to trim both read1 and read2 (if PE) by …  in main()
61 …cmd.add("detect_adapter_for_pe", 0, "by default, the auto-detection for adapter is for SE data inp…  in main()
64 …cmd.add<int>("trim_front1", 'f', "trimming how many bases in front for read1, default is 0", false…  in main()
65 …cmd.add<int>("trim_tail1", 't', "trimming how many bases in tail for read1, default is 0", false, …  in main()
66 …cmd.add<int>("max_len1", 'b', "if read1 is longer than max_len1, then trim read1 at its tail to ma…  in main()
67 …cmd.add<int>("trim_front2", 'F', "trimming how many bases in front for read2. If it's not specifie…  in main()
68 …cmd.add<int>("trim_tail2", 'T', "trimming how many bases in tail for read2. If it's not specified,…  in main()
69 …cmd.add<int>("max_len2", 'B', "if read2 is longer than max_len2, then trim read2 at its tail to ma…  in main()
72     cmd.add("dedup", 'D', "enable deduplication to drop the duplicated reads/pairs");  in main()
73 …cmd.add<int>("dup_calc_accuracy", 0, "accuracy level to calculate duplication (1~6), higher level …  in main()
74 …cmd.add("dont_eval_duplication", 0, "don't evaluate duplication rate to save time and use less mem…  in main()
77 …cmd.add("trim_poly_g", 'g', "force polyG tail trimming, by default trimming is automatically enabl…  in main()
78 …cmd.add<int>("poly_g_min_len", 0, "the minimum length to detect polyG in the read tail. 10 by defa…  in main()
79 …cmd.add("disable_trim_poly_g", 'G', "disable polyG tail trimming, by default trimming is automatic…  in main()
82     cmd.add("trim_poly_x", 'x', "enable polyX trimming in 3' ends.");  in main()
83 …cmd.add<int>("poly_x_min_len", 0, "the minimum length to detect polyX in the read tail. 10 by defa…  in main()
86 …cmd.add("cut_front", '5', "move a sliding window from front (5') to tail, drop the bases in the wi…  in main()
87 …cmd.add("cut_tail", '3', "move a sliding window from tail (3') to front, drop the bases in the win…  in main()
88 …cmd.add("cut_right", 'r', "move a sliding window from front to tail, if meet one window with mean …  in main()
89 …cmd.add<int>("cut_window_size", 'W', "the window size option shared by cut_front, cut_tail or cut_…  in main()
90 …cmd.add<int>("cut_mean_quality", 'M', "the mean quality requirement option shared by cut_front, cu…  in main()
91 …cmd.add<int>("cut_front_window_size", 0, "the window size option of cut_front, default to cut_wind…  in main()
92 …cmd.add<int>("cut_front_mean_quality", 0, "the mean quality requirement option for cut_front, defa…  in main()
93 …cmd.add<int>("cut_tail_window_size", 0, "the window size option of cut_tail, default to cut_window…  in main()
94 …cmd.add<int>("cut_tail_mean_quality", 0, "the mean quality requirement option for cut_tail, defaul…  in main()
95 …cmd.add<int>("cut_right_window_size", 0, "the window size option of cut_right, default to cut_wind…  in main()
96 …cmd.add<int>("cut_right_mean_quality", 0, "the mean quality requirement option for cut_right, defa…  in main()
100 …cmd.add("disable_quality_filtering", 'Q', "quality filtering is enabled by default. If this option…  in main()
101 …cmd.add<int>("qualified_quality_phred", 'q', "the quality value that a base is qualified. Default …  in main()
102 …cmd.add<int>("unqualified_percent_limit", 'u', "how many percents of bases are allowed to be unqua…  in main()
103 …cmd.add<int>("n_base_limit", 'n', "if one read's number of N base is >n_base_limit, then this read…  in main()
104 …cmd.add<int>("average_qual", 'e', "if one read's average quality score <avg_qual, then this read/p…  in main()
107 …cmd.add("disable_length_filtering", 'L', "length filtering is enabled by default. If this option i…  in main()
108 …cmd.add<int>("length_required", 'l', "reads shorter than length_required will be discarded, defaul…  in main()
109 …cmd.add<int>("length_limit", 0, "reads longer than length_limit will be discarded, default 0 means…  in main()
112 …cmd.add("low_complexity_filter", 'y', "enable low complexity filter. The complexity is defined as …  in main()
113 …cmd.add<int>("complexity_threshold", 'Y', "the threshold for low complexity filter (0~100). Defaul…  in main()
116 …cmd.add<string>("filter_by_index1", 0, "specify a file contains a list of barcodes of index1 to be…  in main()
117 …cmd.add<string>("filter_by_index2", 0, "specify a file contains a list of barcodes of index2 to be…  in main()
118 …cmd.add<int>("filter_by_index_threshold", 0, "the allowed difference of index barcode for index fi…  in main()
121 …cmd.add("correction", 'c', "enable base correction in overlapped regions (only for PE data), defau…  in main()
122 …cmd.add<int>("overlap_len_require", 0, "the minimum length to detect overlapped region of PE reads…  in main()
123 …cmd.add<int>("overlap_diff_limit", 0, "the maximum number of mismatched bases to detect overlapped…  in main()
124 …cmd.add<int>("overlap_diff_percent_limit", 0, "the maximum percentage of mismatched bases to detec…  in main()
127     cmd.add("umi", 'U', "enable unique molecular identifier (UMI) preprocessing");  in main()
128 …cmd.add<string>("umi_loc", 0, "specify the location of UMI, can be (index1/index2/read1/read2/per_…  in main()
129 …cmd.add<int>("umi_len", 0, "if the UMI is in read1/read2, its length should be provided", false, 0…  in main()
130 …cmd.add<string>("umi_prefix", 0, "if specified, an underline will be used to connect prefix and UM…  in main()
131 …cmd.add<int>("umi_skip", 0, "if the UMI is in read1/read2, fastp can skip several bases following …  in main()
134     cmd.add("overrepresentation_analysis", 'p', "enable overrepresented sequence analysis.");  in main()
135 …cmd.add<int>("overrepresentation_sampling", 'P', "one in (--overrepresentation_sampling) reads wil…  in main()
138     cmd.add<string>("json", 'j', "the json format report file name", false, "fastp.json");  in main()
139     cmd.add<string>("html", 'h', "the html format report file name", false, "fastp.html");  in main()
140 …cmd.add<string>("report_title", 'R', "should be quoted with \' or \", default is \"fastp report\""…  in main()
143     cmd.add<int>("thread", 'w', "worker thread number, default is 3", false, 3);  in main()
146 …cmd.add<int>("split", 's', "split output by limiting total split file number with this option (2~9…  in main()
147 …cmd.add<long>("split_by_lines", 'S', "split output by limiting lines of each file with this option…  in main()
148 …cmd.add<int>("split_prefix_digits", 'd', "the digits for the sequential number padding (1~10), def…  in main()
151     cmd.add("cut_by_quality5", 0, "DEPRECATED, use --cut_front instead.");  in main()
152     cmd.add("cut_by_quality3", 0, "DEPRECATED, use --cut_tail instead.");  in main()
153     cmd.add("cut_by_quality_aggressive", 0, "DEPRECATED, use --cut_right instead.");  in main()
154     cmd.add("discard_unmerged", 0, "DEPRECATED, no effect now, see the introduction for merging.");  in main()
156     cmd.parse_check(argc, argv);  in main()
159         cerr << cmd.usage() <<endl;  in main()
163     if(cmd.exist("discard_unmerged")) {  in main()
170     opt.in1 = cmd.get<string>("in1");  in main()
171     opt.in2 = cmd.get<string>("in2");  in main()
172     opt.out1 = cmd.get<string>("out1");  in main()
173     opt.out2 = cmd.get<string>("out2");  in main()
174     opt.unpaired1 = cmd.get<string>("unpaired1");  in main()
175     opt.unpaired2 = cmd.get<string>("unpaired2");  in main()
176     opt.failedOut = cmd.get<string>("failed_out");  in main()
177     opt.overlappedOut = cmd.get<string>("overlapped_out");  in main()
181     opt.compression = cmd.get<int>("compression");  in main()
182     opt.readsToProcess = cmd.get<int>("reads_to_process");  in main()
183     opt.phred64 = cmd.exist("phred64");  in main()
184     opt.dontOverwrite = cmd.exist("dont_overwrite");  in main()
185     opt.inputFromSTDIN = cmd.exist("stdin");  in main()
186     opt.outputToSTDOUT = cmd.exist("stdout");  in main()
187     opt.interleavedInput = cmd.exist("interleaved_in");  in main()
188     opt.verbose = cmd.exist("verbose");  in main()
189     opt.fixMGI = cmd.exist("fix_mgi_id");  in main()
192     opt.duplicate.dedup = cmd.exist("dedup");  in main()
193     opt.duplicate.enabled = !cmd.exist("dont_eval_duplication") || cmd.exist("dedup") ;  in main()
194     if(!cmd.exist("dup_calc_accuracy")) {  in main()
200         opt.duplicate.accuracyLevel = min(6, max(1, cmd.get<int>("dup_calc_accuracy")));  in main()
204     opt.merge.enabled = cmd.exist("merge");  in main()
205     opt.merge.out = cmd.get<string>("merged_out");  in main()
206     opt.merge.includeUnmerged = cmd.exist("include_unmerged");  in main()
209     opt.adapter.enabled = !cmd.exist("disable_adapter_trimming");  in main()
210     opt.adapter.detectAdapterForPE = cmd.exist("detect_adapter_for_pe");  in main()
211     opt.adapter.sequence = cmd.get<string>("adapter_sequence");  in main()
212     opt.adapter.sequenceR2 = cmd.get<string>("adapter_sequence_r2");  in main()
213     opt.adapter.fastaFile = cmd.get<string>("adapter_fasta");  in main()
222     opt.trim.front1 = cmd.get<int>("trim_front1");  in main()
223     opt.trim.tail1 = cmd.get<int>("trim_tail1");  in main()
224     opt.trim.maxLen1 = cmd.get<int>("max_len1");  in main()
226     if(cmd.exist("trim_front2"))  in main()
227         opt.trim.front2 = cmd.get<int>("trim_front2");  in main()
230     if(cmd.exist("trim_tail2"))  in main()
231         opt.trim.tail2 = cmd.get<int>("trim_tail2");  in main()
234     if(cmd.exist("max_len2"))  in main()
235         opt.trim.maxLen2 = cmd.get<int>("max_len2");  in main()
240     if(cmd.exist("trim_poly_g") && cmd.exist("disable_trim_poly_g")) {  in main()
242     } else if(cmd.exist("trim_poly_g")) {  in main()
244     } else if(cmd.exist("disable_trim_poly_g")) {  in main()
247     opt.polyGTrim.minLen = cmd.get<int>("poly_g_min_len");  in main()
250     if(cmd.exist("trim_poly_x")) {  in main()
253     opt.polyXTrim.minLen = cmd.get<int>("poly_x_min_len");  in main()
257     opt.qualityCut.enabledFront = cmd.exist("cut_front");  in main()
260         opt.qualityCut.enabledFront = cmd.exist("cut_by_quality5");  in main()
264     opt.qualityCut.enabledTail = cmd.exist("cut_tail");  in main()
267         opt.qualityCut.enabledFront = cmd.exist("cut_by_quality3");  in main()
271     opt.qualityCut.enabledRight = cmd.exist("cut_right");  in main()
274         opt.qualityCut.enabledRight = cmd.exist("cut_by_quality_aggressive");  in main()
279     opt.qualityCut.windowSizeShared = cmd.get<int>("cut_window_size");  in main()
280     opt.qualityCut.qualityShared = cmd.get<int>("cut_mean_quality");  in main()
282     if(cmd.exist("cut_front_window_size"))  in main()
283         opt.qualityCut.windowSizeFront = cmd.get<int>("cut_front_window_size");  in main()
286     if(cmd.exist("cut_front_mean_quality"))  in main()
287         opt.qualityCut.qualityFront = cmd.get<int>("cut_front_mean_quality");  in main()
291     if(cmd.exist("cut_tail_window_size"))  in main()
292         opt.qualityCut.windowSizeTail = cmd.get<int>("cut_tail_window_size");  in main()
295     if(cmd.exist("cut_tail_mean_quality"))  in main()
296         opt.qualityCut.qualityTail = cmd.get<int>("cut_tail_mean_quality");  in main()
300     if(cmd.exist("cut_right_window_size"))  in main()
301         opt.qualityCut.windowSizeRight = cmd.get<int>("cut_right_window_size");  in main()
304     if(cmd.exist("cut_right_mean_quality"))  in main()
305         opt.qualityCut.qualityRight = cmd.get<int>("cut_right_mean_quality");  in main()
311         if(cmd.exist("cut_window_size") || cmd.exist("cut_mean_quality")   in main()
312             || cmd.exist("cut_front_window_size") || cmd.exist("cut_front_mean_quality")   in main()
313             || cmd.exist("cut_tail_window_size") || cmd.exist("cut_tail_mean_quality")   in main()
314             || cmd.exist("cut_right_window_size") || cmd.exist("cut_right_mean_quality"))  in main()
319     opt.qualfilter.enabled = !cmd.exist("disable_quality_filtering");  in main()
320     opt.qualfilter.qualifiedQual = num2qual(cmd.get<int>("qualified_quality_phred"));  in main()
321     opt.qualfilter.unqualifiedPercentLimit = cmd.get<int>("unqualified_percent_limit");  in main()
322     opt.qualfilter.avgQualReq = cmd.get<int>("average_qual");  in main()
323     opt.qualfilter.nBaseLimit = cmd.get<int>("n_base_limit");  in main()
326     opt.lengthFilter.enabled = !cmd.exist("disable_length_filtering");  in main()
327     opt.lengthFilter.requiredLength = cmd.get<int>("length_required");  in main()
328     opt.lengthFilter.maxLength = cmd.get<int>("length_limit");  in main()
331     opt.complexityFilter.enabled = cmd.exist("low_complexity_filter");  in main()
332 … opt.complexityFilter.threshold = (min(100, max(0, cmd.get<int>("complexity_threshold")))) / 100.0;  in main()
335     opt.correction.enabled = cmd.exist("correction");  in main()
336     opt.overlapRequire = cmd.get<int>("overlap_len_require");  in main()
337     opt.overlapDiffLimit = cmd.get<int>("overlap_diff_limit");  in main()
338     opt.overlapDiffPercentLimit = cmd.get<int>("overlap_diff_percent_limit");  in main()
341     opt.thread = cmd.get<int>("thread");  in main()
344     opt.jsonFile = cmd.get<string>("json");  in main()
345     opt.htmlFile = cmd.get<string>("html");  in main()
346     opt.reportTitle = cmd.get<string>("report_title");  in main()
349     opt.split.enabled = cmd.exist("split") || cmd.exist("split_by_lines");  in main()
350     opt.split.digits = cmd.get<int>("split_prefix_digits");  in main()
351     if(cmd.exist("split") && cmd.exist("split_by_lines")) {  in main()
354     if(cmd.exist("split")) {  in main()
355         opt.split.number = cmd.get<int>("split");  in main()
359     if(cmd.exist("split_by_lines")) {  in main()
360         long lines = cmd.get<long>("split_by_lines");  in main()
376     opt.umi.enabled = cmd.exist("umi");  in main()
377     opt.umi.length = cmd.get<int>("umi_len");  in main()
378     opt.umi.prefix = cmd.get<string>("umi_prefix");  in main()
379     opt.umi.skip = cmd.get<int>("umi_skip");  in main()
381         string umiLoc = cmd.get<string>("umi_loc");  in main()
408     opt.overRepAnalysis.enabled = cmd.exist("overrepresentation_analysis");  in main()
409     opt.overRepAnalysis.sampling = cmd.get<int>("overrepresentation_sampling");  in main()
412     string blacklist1 = cmd.get<string>("filter_by_index1");  in main()
413     string blacklist2 = cmd.get<string>("filter_by_index2");  in main()
414     int indexFilterThreshold = cmd.get<int>("filter_by_index_threshold");  in main()
492     if(!cmd.exist("trim_poly_g") && !cmd.exist("disable_trim_poly_g") && supportEvaluation) {  in main()