Mercurial > repos > siyuan > prada

comparison pyPRADA_1.2/prada-preprocess-bi @ 0:acc2ca1a3ba4

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
Uploaded
author siyuan
date Thu, 20 Feb 2014 00:44:58 -0500
parents
children
comparison
equal deleted inserted replaced
-1:000000000000 0:acc2ca1a3ba4
1 #!/usr/bin/env python
2
3 #This is a program to implement the PRADA pipeline "process subsection", based on the work of Rahul Vegesna and Wandaliz Torres-Garcia
4 #It merely generates a PBS file, depending on the user input entry point (step)
5 #Author: Siyuan Zheng, szheng2@mdanderson.org
6 #Copy right belongs to Roel Verhaak's lab from MD Anderson Cancer Center, Department of Bioinformatics and Computational Biology.
7 #Last revision: 02/17/2014
8
9 import subprocess
10 import os,os.path
11 import sys
12 import time
13 import ioprada
14 import re
15
16 ########################################################################################
17 args=sys.argv
18
19 help_menu='''\nPipeline for RNAseq Data Analaysis - preprocessing pipeline (PRADA).
20 \t**Usage**:
21 \tprada-preprocess-bi -conf xx.txt -inputdir .. -sample XX -tag TCGA-XX -platform illumina -step 1_1 -intermediate no -pbs xxx -outdir ... -submit no
22 \t**Parameters**:
23 \t-h print help message
24 \t-step_info print complete steps curated in the module.
25 \t-inputdir the dir where the input bam or fastq can be found.
26 \t-sample input sample name. PRADA searches for sample.bam or sample.end1.fastq/sample.end2.fastq, etc, depending on the
27 \t initiating step number. See step_info for more information.
28 \t-conf config file for references and parameters. Default is conf.txt in py-PRADA installation folder.
29 \t-tag a tag to describe the sample, likely sample ID, such as TCGA-LGG-01; no default.
30 \t-platform only illumina at present (default).
31 \t-step values: 1_1/2,2_e1/2_1/2/3/4,3_e1/2_1/2,4_1/2,5,6_1/2,7,8; example 2_e1_1; no default.
32 \t-outdir output dir. Default is the directory where the input bam is.
33 \t-pbs name for output pbs file and log file. Default (time-stamp) is used if no input.
34 \t-intermediate values:yes/no; if intermediate files should be kept. Default is not.
35 \t-submit if submit the job to HPC, default is no. If yes, ppn is set to 12.
36 \t-v print version information.
37 '''
38
39 steps_info='''
40 Command orders (sample XX)
41 step 1_1 --> XX.sorted.bam [sort input bam by name]
42 step 1_2 --> XX.end1/2.fastq [extract reads from bam]
43 step 2_e1_1 --> XX.end1.sai [realign end1 reads to composite reference]
44 step 2_e1_2 --> XX.end1.sam [generate sam]
45 step 2_e1_3 --> XX.end1.bam [generate bam]
46 step 2_e1_4 --> XX.end1.sorted.bam [sort bam]
47 step 2_e2_1 --> XX.end2.sai [realign end2 reads to composite reference]
48 step 2_e2_2 --> XX.end2.sam [generate sam]
49 step 2_e2_3 --> XX.end2.bam [generate bam]
50 step 2_e2_4 --> XX.end2.sorted.bam [sort bam]
51 step 3_e1_1 --> XX.end1.remapped.bam [remap end1 to genome]
52 step 3_e1_2 --> XX.end1.remapped.sorted.bam [sort remapped bam]
53 step 3_e2_1 --> XX.end2.remapped.bam [remap end2 to genome]
54 step 3_e2_2 --> XX.end2.remapped.sorted.bam [sort remapped bam]
55 step 4_1 --> XX.paired.bam [pair end1 and end1]
56 step 4_2 --> XX.paired.sorted.bam [sort paired bam]
57 step 5 --> XX.withRG.paired.sorted.bam [add read group]
58 step 6_1 --> XX.orig.csv [prepair recalibration table]
59 step 6_2 --> XX.withRG.GATKRecalibrated.bam [recalibration]
60 step 7 --> XX.withRG.GATKRecalibrated.flagged.bam [flag duplication reads]
61 step 8 --> folder XX for gene expression, QC metrics etc. [generate QC and expression]
62 '''
63
64 if '-h' in args or '-help' in args or len(args)==1:
65 print help_menu
66 sys.exit(0)
67 if '-step_info' in args:
68 print steps_info
69 sys.exit(0)
70 if '-v' in args:
71 import version
72 print version.version
73 sys.exit(0)
74
75 if '-sample' not in args:
76 sys.exit('ERROR: Sample name is needed')
77 if '-step' not in args:
78 sys.exit('ERROR: Step number is needed')
79 if '-tag' not in args:
80 sys.exit('ERROR: A tag is needed')
81
82 i=args.index('-sample')
83 sample=args[i+1]
84 if '-inputdir' not in args:
85 inputpath=os.path.abspath('./')
86 else:
87 i=args.index('-inputdir')
88 inputpath=args[i+1]
89 bampath=os.path.abspath(inputpath)+'/%s.bam'%sample
90 fq1path=os.path.abspath(inputpath)+'/%s.end1.fastq'%sample
91 fq2path=os.path.abspath(inputpath)+'/%s.end2.fastq'%sample
92
93 if '-outdir' not in args:
94 outpath=os.path.dirname(bampath)
95 else:
96 i=args.index('-outdir')
97 outpath=os.path.abspath(args[i+1])
98 if not os.path.exists(outpath):
99 os.mkdir(outpath)
100
101 #bam=os.path.basename(bampath)
102 #sample=bam[:-4]
103 i=args.index('-step')
104 step=args[i+1]
105 i=args.index('-tag')
106 tag=args[i+1]
107
108 prada_path=os.path.dirname(os.path.abspath(__file__)) ####
109 ref_search_path=[prada_path,os.getcwd()] #search path for ref file if not specified in command line
110
111 if '-conf' in args:
112 i=args.index('-conf')
113 reffile=args[i+1]
114 if os.path.exists(reffile):
115 pass
116 else:
117 for pth in ref_search_path:
118 new_reffile='%s/%s'%(pth, os.path.basename(reffile))
119 if os.path.exists(new_reffile):
120 reffile=new_reffile
121 break
122 else:
123 sys.exit('ERROR: conf file %s not found'%reffile)
124 else:
125 reffile='%s/conf.txt'%prada_path
126 if not os.path.exists(reffile):
127 sys.exit('ERROR: No default conf.txt found and none specified')
128
129 if '-platform' in args:
130 i=args.index('-platform')
131 plat=args[i+1]
132 else:
133 plat='illumina'
134 if '-submit' in args:
135 i=args.index('-submit')
136 submit=args[i+1]
137 else:
138 submit='no'
139 if '-intermediate' in args:
140 i=args.index('-intermediate')
141 keepmed=args[i+1]
142 else:
143 keepmed='False'
144
145 if keepmed in ['False','FALSE','false','F','NO','No','no','n']:
146 keepmed_flag='no'
147 elif keepmed in ['True','TRUE','true','T','YES','Yes','yes','y']:
148 keepmed_flag='yes'
149 else:
150 sys.exit('ERROR: -intermediate value not recognized')
151
152 if '-pbs' in args:
153 i=args.index('-pbs')
154 docstr=args[i+1]
155 else:
156 a=time.ctime().split()
157 b=time.time()
158 timestamp='_'.join([a[-1],a[1],a[2]])+'.'+str(b)
159 docstr='prada_prep_'+timestamp
160 logfilename=docstr+'.log'
161 pbsfilename=docstr+'.pbs'
162 pbspath=outpath+'/'+pbsfilename
163 logpath=outpath+'/'+logfilename
164 ########################################################################################
165
166 ########################################################################################
167 #underlying utilities, automatically detected
168 samtools='%s/tools/samtools-0.1.16/samtools'%prada_path
169 bwa='%s/tools/bwa-0.5.7-mh/bwa'%prada_path
170 gatk='%s/tools/GATK/'%prada_path
171 picard='%s/tools/Picard/'%prada_path
172 seqc='%s/tools/RNA-SeQC_v1.1.7.jar'%prada_path
173 #Default uses 12 nodes in HPC
174 #########################################################################################
175
176 #########################################################################################
177 #reference files
178 refdict=ioprada.read_conf(reffile)
179 genome_gtf=refdict['--REF--']['genome_gtf']
180 compdb_fasta=refdict['--REF--']['compdb_fasta']
181 compdb_map=refdict['--REF--']['compdb_map']
182 genome_fasta=refdict['--REF--']['genome_fasta']
183 dbsnp_vcf=refdict['--REF--']['dbsnp_vcf']
184 select_tx=refdict['--REF--']['select_tx']
185 pat=re.compile('ppn=(\d*)')
186 parallel_n=pat.search(refdict['--PBS--']['-l']).groups()[0]
187 #########################################################################################
188
189 #########################################################################################
190 #pipeline command lines.
191 ##Cleaning up steps: if -intermediate is yes, none will be executed.
192 step_1_1_cmd=['%s sort -n -m 1000000000 %s %s.sorted'%(samtools,bampath,sample)]
193 post_1_1_clean=[]
194 step_1_2_cmd=['java -Xmx8g -jar %s/SamToFastq.jar INPUT=%s.sorted.bam FASTQ=%s.end1.fastq SECOND_END_FASTQ=%s.end2.fastq INCLUDE_NON_PF_READS=true VALIDATION_STRINGENCY=SILENT TMP_DIR=tmp/'%(picard,sample,sample,sample)]
195 post_1_2_clean=['rm -f %s.sorted.bam'%sample]
196 alnparams=' '.join([' '.join(x) for x in refdict['--BWA aln--'].items()])
197 samseparams=' '.join([' '.join(x) for x in refdict['--BWA samse--'].items()])
198 if step == '2_e1_1' or step == '2_e2_1':
199 step_2_e1_1_cmd=['%s aln %s %s %s > %s.end1.sai'%(bwa,alnparams,compdb_fasta,fq1path,sample)]
200 post_2_e1_1_clean=[]
201 step_2_e1_2_cmd=['%s samse -s %s %s %s.end1.sai %s > %s.end1.sam'%(bwa,samseparams,compdb_fasta,sample,fq1path,sample)]
202 post_2_e1_2_clean=['rm -f %s.end1.sai'%sample]
203 step_2_e2_1_cmd=['%s aln %s %s %s > %s.end2.sai'%(bwa,alnparams,compdb_fasta,fq2path,sample)]
204 post_2_e2_1_clean=[]
205 step_2_e2_2_cmd=['%s samse -s %s %s %s.end2.sai %s > %s.end2.sam'%(bwa,samseparams,compdb_fasta,sample,fq2path,sample)]
206 post_2_e2_2_clean=['rm -f %s.end2.sai'%sample]
207 else:
208 step_2_e1_1_cmd=['%s aln %s %s %s.end1.fastq > %s.end1.sai'%(bwa,alnparams,compdb_fasta,sample,sample)]
209 post_2_e1_1_clean=[]
210 step_2_e1_2_cmd=['%s samse -s %s %s %s.end1.sai %s.end1.fastq > %s.end1.sam'%(bwa,samseparams,compdb_fasta,sample,sample,sample)]
211 post_2_e1_2_clean=['rm -f %s.end1.sai'%sample,'rm -f %s.end1.fastq'%sample]
212 step_2_e2_1_cmd=['%s aln %s %s %s.end2.fastq > %s.end2.sai'%(bwa,alnparams,compdb_fasta,sample,sample)]
213 post_2_e2_1_clean=[]
214 step_2_e2_2_cmd=['%s samse -s %s %s %s.end2.sai %s.end2.fastq > %s.end2.sam'%(bwa,samseparams,compdb_fasta,sample,sample,sample)]
215 post_2_e2_2_clean=['rm -f %s.end2.sai'%sample,'rm -f %s.end2.fastq'%sample]
216 step_2_e1_3_cmd=['%s view -bS -o %s.end1.bam %s.end1.sam'%(samtools,sample,sample)]
217 post_2_e1_3_clean=['rm -f %s.end1.sam'%sample]
218 step_2_e1_4_cmd=['%s sort -n -m 1000000000 %s.end1.bam %s.end1.sorted'%(samtools,sample,sample)]
219 post_2_e1_4_clean=['rm -f %s.end1.bam'%sample]
220 step_2_e2_3_cmd=['%s view -bS -o %s.end2.bam %s.end2.sam'%(samtools,sample,sample)]
221 post_2_e2_3_clean=['rm -f %s.end2.sam'%sample]
222 step_2_e2_4_cmd=['%s sort -n -m 1000000000 %s.end2.bam %s.end2.sorted'%(samtools,sample,sample)]
223 post_2_e2_4_clean=['rm -f %s.end2.bam'%sample]
224 step_3_e1_1_cmd=['java -Djava.io.tmpdir=tmp/ -cp %s/RemapAlignments.jar -Xmx8g org.broadinstitute.cga.tools.gatk.rna.RemapAlignments M=%s IN=%s.end1.sorted.bam OUT=%s.end1.remapped.bam R=%s REDUCE=TRUE'%(gatk,compdb_map,sample,sample,genome_fasta)]
225 post_3_e1_1_clean=['rm -f %s.end1.sorted.bam'%sample]
226 step_3_e1_2_cmd=['%s sort -n -m 1000000000 %s.end1.remapped.bam %s.end1.remapped.sorted'%(samtools,sample,sample)]
227 post_3_e1_2_clean=['rm -f %s.end1.remapped.bam'%sample]
228 step_3_e2_1_cmd=['java -Djava.io.tmpdir=tmp/ -cp %s/RemapAlignments.jar -Xmx8g org.broadinstitute.cga.tools.gatk.rna.RemapAlignments M=%s IN=%s.end2.sorted.bam OUT=%s.end2.remapped.bam R=%s REDUCE=TRUE'%(gatk,compdb_map,sample,sample,genome_fasta)]
229 post_3_e2_1_clean=['rm -f %s.end2.sorted.bam'%sample]
230 step_3_e2_2_cmd=['%s sort -n -m 1000000000 %s.end2.remapped.bam %s.end2.remapped.sorted'%(samtools,sample,sample)]
231 post_3_e2_2_clean=['rm -f %s.end2.remapped.bam'%sample]
232 step_4_1_cmd=['java -Djava.io.tmpdir=tmp/ -Xmx8g -jar %s/PairMaker.jar IN1=%s.end1.remapped.sorted.bam IN2=%s.end2.remapped.sorted.bam OUTPUT=%s.paired.bam TMP_DIR=tmp/'%(gatk,sample,sample,sample)]
233 post_4_1_clean=['rm -f %s.end1.remapped.sorted.bam'%sample,'rm -f %s.end2.remapped.sorted.bam'%sample]
234 step_4_2_cmd=['%s sort -m 1000000000 %s.paired.bam %s.paired.sorted'%(samtools,sample,sample)]
235 post_4_2_clean=['rm -f %s.paired.bam'%sample]
236 step_5_cmd=['java -Xmx8g -jar %s/AddOrReplaceReadGroups.jar I=%s.paired.sorted.bam O=%s.withRG.paired.sorted.bam RGLB=%s RGPL=%s RGPU=%s RGSM=%s'%(picard,sample,sample,tag,plat,tag,tag)]
237 post_5_clean=['rm -f %s.paired.sorted.bam'%sample]
238 step_6_1_cmd=['%s index %s.withRG.paired.sorted.bam'%(samtools,sample),'java -Xmx8g -jar %s/GenomeAnalysisTK.jar -l INFO -R %s --default_platform %s --knownSites %s -I %s.withRG.paired.sorted.bam --downsample_to_coverage 10000 -T CountCovariates -cov ReadGroupCovariate -cov QualityScoreCovariate -cov CycleCovariate -cov DinucCovariate -nt %s -recalFile %s.orig.csv'%(gatk,genome_fasta,plat,dbsnp_vcf,sample,parallel_n,sample)]
239 post_6_1_clean=[]
240 step_6_2_cmd=['java -Xmx8g -jar %s/GenomeAnalysisTK.jar -l INFO -R %s --default_platform %s -I %s.withRG.paired.sorted.bam -T TableRecalibration --out %s.withRG.GATKRecalibrated.bam -recalFile %s.orig.csv'%(gatk,genome_fasta,plat,sample,sample,sample)]
241 post_6_2_clean=['rm -f %s.withRG.paired.sorted.bam'%sample,'rm -f %s.withRG.paired.sorted.bam.bai'%sample,'rm -f %s.orig.csv'%sample]
242 step_7_cmd=['java -Xmx8g -jar %s/MarkDuplicates.jar I=%s.withRG.GATKRecalibrated.bam O=%s.withRG.GATKRecalibrated.flagged.bam METRICS_FILE=%s.Duplicates_metrics.txt VALIDATION_STRINGENCY=SILENT TMP_DIR=tmp/'%(picard,sample,sample,sample),'%s index %s.withRG.GATKRecalibrated.flagged.bam'%(samtools,sample)]
243 post_7_clean=['rm -f %s.withRG.GATKRecalibrated.bam'%sample,'rm -f %s.withRG.GATKRecalibrated.bai'%sample,'rm -f %s.Duplicates_metrics.txt'%sample]
244 step_8_cmd=["java -Xmx8g -jar %s -ttype 2 -t %s -r %s -s '%s|%s.withRG.GATKRecalibrated.flagged.bam|Disc' -o %s/"%(seqc,genome_gtf,genome_fasta,sample,sample,sample)]
245 post_8_clean=[]
246
247 cmdset=[]
248 for item in globals().keys():
249 if item.endswith('_cmd'):
250 cmdset.append(item)
251 cmdset.sort()
252
253 #########################################################################################
254 #write PBS file
255
256 ###############Entry Point
257 try:
258 cmd_entry=cmdset.index('step_'+step+'_cmd') ##entry point
259 except ValueError:
260 sys.exit('ERROR: STEP not recognized')
261
262 def _parsecmd(cmd):
263 '''pase cmd str for step information'''
264 info=cmd.split('_')
265 cstep='_'.join(info[1:-1])
266 return cstep
267
268 ###########################
269 ##headers
270 outfile=open(pbspath,'w')
271 outfile.write('#! /bin/sh\n')
272 outfile.write('#PBS -V\n')
273 outfile.write('#PBS -N %s\n'%tag)
274 outfile.write('#PBS -j oe\n')
275 outfile.write('#PBS -o %s\n'%logpath)
276 for item in refdict['--PBS--']:
277 outfile.write('#PBS %s %s\n'%(item,refdict['--PBS--'][item]))
278 outfile.write('#PBS -d %s\n'%outpath)
279
280 #commands
281 outfile.write('echo "Job start: `date`"\n')
282
283 for i in range(cmd_entry,len(cmdset)):
284 cmd=cmdset[i]
285 cstep=_parsecmd(cmd)
286 clean_flag=1
287 outfile.write('echo "step %s start: `date`"\n'%cstep)
288 outfile.write('if\n')
289 outfile.write('\t%s\n'%('\n'.join(eval(cmd))))
290 outfile.write('then\n')
291 outfile.write('\techo "step %s done: `date`"\n'%cstep)
292 outfile.write('else\n')
293 outfile.write('\techo "step %s ERROR"\n'%cstep)
294 outfile.write('\texit\n')
295 outfile.write('fi\n')
296
297 if keepmed_flag=='yes': #if so, keep files by force
298 clean_flag=0
299
300 if clean_flag==1:
301 outfile.write('\n'.join(eval('post_%s_clean'%cstep)))
302 outfile.write('\n')
303
304 outfile.write('echo "PIPELINE FINISHED"\n')
305 outfile.close()
306 ######################################################################
307
308 if submit in ['False','FALSE','false','F','NO','No','no','n']:
309 jid='Not_Submitted'
310 logpath='None'
311 elif submit in ['True','TRUE','true','T','YES','Yes','yes','y']:
312 cmdstr='qsub %s'%pbspath
313 cmd=cmdstr.split()
314 cmdout=subprocess.Popen(cmd,stdout=subprocess.PIPE,stderr=subprocess.STDOUT)
315 jid=cmdout.stdout.read().strip() ##JOB ID
316 else:
317 sys.exit('ERROR: submit parameter not recognized')
318
319 print '#!#%s'%tag
320 print 'BAM\t%s'%bampath
321 print 'Entry\t%s'%step
322 print 'PBS\t%s'%pbspath
323 print 'JOB\t%s'%jid
324 print 'LOG\t%s'%logpath
325