#!/usr/bin/env python

#This is a program to implement the PRADA pipeline "process subsection", based on the work of Rahul Vegesna and Wandaliz Torres-Garcia
#It merely generates a PBS file, depending on the user input entry point (step) 
#Author: Siyuan Zheng, szheng2@mdanderson.org
#Copy right belongs to Roel Verhaak's lab from MD Anderson Cancer Center, Department of Bioinformatics and Computational Biology. 
#Last revision: 02/17/2014

import subprocess
import os,os.path
import sys
import time
import ioprada
import re

########################################################################################
args=sys.argv

help_menu='''\nPipeline for RNAseq Data Analaysis - preprocessing pipeline (PRADA).
\t**Usage**:
\tprada-preprocess-bi -conf xx.txt -inputdir .. -sample XX -tag TCGA-XX -platform illumina -step 1_1 -intermediate no -pbs xxx -outdir ... -submit no
\t**Parameters**:
\t-h		print help message
\t-step_info	print complete steps curated in the module.
\t-inputdir	the dir where the input bam or fastq can be found.
\t-sample		input sample name. PRADA searches for sample.bam or sample.end1.fastq/sample.end2.fastq, etc, depending on the 
\t		initiating step number. See step_info for more information. 
\t-conf		config file for references and parameters. Default is conf.txt in py-PRADA installation folder. 
\t-tag		a tag to describe the sample, likely sample ID, such as TCGA-LGG-01; no default.
\t-platform	only illumina at present (default). 
\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.
\t-outdir		output dir. Default is the directory where the input bam is. 
\t-pbs		name for output pbs file and log file. Default (time-stamp) is used if no input. 
\t-intermediate	values:yes/no; if intermediate files should be kept. Default is not. 
\t-submit		if submit the job to HPC, default is no. If yes, ppn is set to 12.
\t-v		print version information. 
'''

steps_info='''
Command orders (sample XX)
step 1_1	-->  XX.sorted.bam [sort input bam by name]
step 1_2	-->  XX.end1/2.fastq [extract reads from bam]
step 2_e1_1	-->  XX.end1.sai [realign end1 reads to composite reference]
step 2_e1_2	-->  XX.end1.sam [generate sam]
step 2_e1_3	-->  XX.end1.bam [generate bam]
step 2_e1_4	-->  XX.end1.sorted.bam [sort bam]
step 2_e2_1	-->  XX.end2.sai [realign end2 reads to composite reference]
step 2_e2_2	-->  XX.end2.sam [generate sam]
step 2_e2_3	-->  XX.end2.bam [generate bam]
step 2_e2_4	-->  XX.end2.sorted.bam [sort bam]
step 3_e1_1	-->  XX.end1.remapped.bam [remap end1 to genome]
step 3_e1_2	-->  XX.end1.remapped.sorted.bam [sort remapped bam]
step 3_e2_1	-->  XX.end2.remapped.bam [remap end2 to genome]
step 3_e2_2	-->  XX.end2.remapped.sorted.bam [sort remapped bam]
step 4_1	-->  XX.paired.bam [pair end1 and end1]
step 4_2	-->  XX.paired.sorted.bam [sort paired bam]
step 5		-->  XX.withRG.paired.sorted.bam [add read group]
step 6_1	-->  XX.orig.csv [prepair recalibration table]
step 6_2	-->  XX.withRG.GATKRecalibrated.bam [recalibration]
step 7		-->  XX.withRG.GATKRecalibrated.flagged.bam [flag duplication reads]
step 8		-->  folder XX for gene expression, QC metrics etc. [generate QC and expression]
'''

if '-h' in args or '-help' in args or len(args)==1:
    print help_menu
    sys.exit(0)
if '-step_info' in args:
    print steps_info
    sys.exit(0)
if '-v' in args:
    import version
    print version.version
    sys.exit(0)

if '-sample' not in args:
    sys.exit('ERROR: Sample name is needed')
if '-step' not in args:
    sys.exit('ERROR: Step number is needed')
if '-tag' not in args:
    sys.exit('ERROR: A tag is needed')

i=args.index('-sample')
sample=args[i+1]
if '-inputdir' not in args:
    inputpath=os.path.abspath('./')
else:
    i=args.index('-inputdir')
    inputpath=args[i+1]
bampath=os.path.abspath(inputpath)+'/%s.bam'%sample
fq1path=os.path.abspath(inputpath)+'/%s.end1.fastq'%sample
fq2path=os.path.abspath(inputpath)+'/%s.end2.fastq'%sample

if '-outdir' not in args:
    outpath=os.path.dirname(bampath)
else:
    i=args.index('-outdir')
    outpath=os.path.abspath(args[i+1])
if not os.path.exists(outpath):
    os.mkdir(outpath)

#bam=os.path.basename(bampath)
#sample=bam[:-4]
i=args.index('-step')
step=args[i+1]
i=args.index('-tag')
tag=args[i+1]

prada_path=os.path.dirname(os.path.abspath(__file__))   ####
ref_search_path=[prada_path,os.getcwd()]                #search path for ref file if not specified in command line

if '-conf' in args:
    i=args.index('-conf')
    reffile=args[i+1]
    if os.path.exists(reffile):
        pass
    else:
        for pth in ref_search_path:
            new_reffile='%s/%s'%(pth, os.path.basename(reffile))
            if os.path.exists(new_reffile):
                reffile=new_reffile
                break
        else:
            sys.exit('ERROR: conf file %s not found'%reffile)
else:
    reffile='%s/conf.txt'%prada_path
    if not os.path.exists(reffile):
        sys.exit('ERROR: No default conf.txt found and none specified')

if '-platform' in args:
    i=args.index('-platform')
    plat=args[i+1]
else:
    plat='illumina'
if '-submit' in args:
    i=args.index('-submit')
    submit=args[i+1]
else:
    submit='no'
if '-intermediate' in args:
    i=args.index('-intermediate')
    keepmed=args[i+1]
else:
    keepmed='False'

if keepmed in ['False','FALSE','false','F','NO','No','no','n']:
    keepmed_flag='no'
elif keepmed in ['True','TRUE','true','T','YES','Yes','yes','y']:
    keepmed_flag='yes'
else:
    sys.exit('ERROR: -intermediate value not recognized')

if '-pbs' in args:
    i=args.index('-pbs')
    docstr=args[i+1]
else:
    a=time.ctime().split()
    b=time.time()
    timestamp='_'.join([a[-1],a[1],a[2]])+'.'+str(b)
    docstr='prada_prep_'+timestamp
logfilename=docstr+'.log'
pbsfilename=docstr+'.pbs'
pbspath=outpath+'/'+pbsfilename
logpath=outpath+'/'+logfilename
########################################################################################

########################################################################################
#underlying utilities, automatically detected
samtools='%s/tools/samtools-0.1.16/samtools'%prada_path
bwa='%s/tools/bwa-0.5.7-mh/bwa'%prada_path
gatk='%s/tools/GATK/'%prada_path
picard='%s/tools/Picard/'%prada_path
seqc='%s/tools/RNA-SeQC_v1.1.7.jar'%prada_path
#Default uses 12 nodes in HPC
#########################################################################################

#########################################################################################
#reference files
refdict=ioprada.read_conf(reffile)
genome_gtf=refdict['--REF--']['genome_gtf']
compdb_fasta=refdict['--REF--']['compdb_fasta']
compdb_map=refdict['--REF--']['compdb_map']
genome_fasta=refdict['--REF--']['genome_fasta']
dbsnp_vcf=refdict['--REF--']['dbsnp_vcf']
select_tx=refdict['--REF--']['select_tx']
pat=re.compile('ppn=(\d*)')
parallel_n=pat.search(refdict['--PBS--']['-l']).groups()[0]
#########################################################################################

#########################################################################################
#pipeline command lines.
##Cleaning up steps: if -intermediate is yes, none will be executed.
step_1_1_cmd=['%s sort -n -m 1000000000 %s %s.sorted'%(samtools,bampath,sample)]
post_1_1_clean=[]
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)]
post_1_2_clean=['rm -f %s.sorted.bam'%sample]
alnparams=' '.join([' '.join(x) for x in refdict['--BWA aln--'].items()])
samseparams=' '.join([' '.join(x) for x in refdict['--BWA samse--'].items()])
if step == '2_e1_1' or step == '2_e2_1':
    step_2_e1_1_cmd=['%s aln %s %s %s > %s.end1.sai'%(bwa,alnparams,compdb_fasta,fq1path,sample)]
    post_2_e1_1_clean=[]
    step_2_e1_2_cmd=['%s samse -s %s %s %s.end1.sai %s > %s.end1.sam'%(bwa,samseparams,compdb_fasta,sample,fq1path,sample)]
    post_2_e1_2_clean=['rm -f %s.end1.sai'%sample]
    step_2_e2_1_cmd=['%s aln %s %s %s > %s.end2.sai'%(bwa,alnparams,compdb_fasta,fq2path,sample)]
    post_2_e2_1_clean=[]
    step_2_e2_2_cmd=['%s samse -s %s %s %s.end2.sai %s > %s.end2.sam'%(bwa,samseparams,compdb_fasta,sample,fq2path,sample)]
    post_2_e2_2_clean=['rm -f %s.end2.sai'%sample]
else:
    step_2_e1_1_cmd=['%s aln %s %s %s.end1.fastq > %s.end1.sai'%(bwa,alnparams,compdb_fasta,sample,sample)]
    post_2_e1_1_clean=[]
    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)]
    post_2_e1_2_clean=['rm -f %s.end1.sai'%sample,'rm -f %s.end1.fastq'%sample]
    step_2_e2_1_cmd=['%s aln %s %s %s.end2.fastq > %s.end2.sai'%(bwa,alnparams,compdb_fasta,sample,sample)]
    post_2_e2_1_clean=[]
    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)]
    post_2_e2_2_clean=['rm -f %s.end2.sai'%sample,'rm -f %s.end2.fastq'%sample]
step_2_e1_3_cmd=['%s view -bS -o %s.end1.bam %s.end1.sam'%(samtools,sample,sample)]
post_2_e1_3_clean=['rm -f %s.end1.sam'%sample]
step_2_e1_4_cmd=['%s sort -n -m 1000000000 %s.end1.bam %s.end1.sorted'%(samtools,sample,sample)]
post_2_e1_4_clean=['rm -f %s.end1.bam'%sample]
step_2_e2_3_cmd=['%s view -bS -o %s.end2.bam %s.end2.sam'%(samtools,sample,sample)]
post_2_e2_3_clean=['rm -f %s.end2.sam'%sample]
step_2_e2_4_cmd=['%s sort -n -m 1000000000 %s.end2.bam %s.end2.sorted'%(samtools,sample,sample)]
post_2_e2_4_clean=['rm -f %s.end2.bam'%sample]
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)]
post_3_e1_1_clean=['rm -f %s.end1.sorted.bam'%sample]
step_3_e1_2_cmd=['%s sort -n -m 1000000000 %s.end1.remapped.bam %s.end1.remapped.sorted'%(samtools,sample,sample)]
post_3_e1_2_clean=['rm -f %s.end1.remapped.bam'%sample]
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)]
post_3_e2_1_clean=['rm -f %s.end2.sorted.bam'%sample]
step_3_e2_2_cmd=['%s sort -n -m 1000000000 %s.end2.remapped.bam %s.end2.remapped.sorted'%(samtools,sample,sample)]
post_3_e2_2_clean=['rm -f %s.end2.remapped.bam'%sample]
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)]
post_4_1_clean=['rm -f %s.end1.remapped.sorted.bam'%sample,'rm -f %s.end2.remapped.sorted.bam'%sample]
step_4_2_cmd=['%s sort -m 1000000000 %s.paired.bam %s.paired.sorted'%(samtools,sample,sample)]
post_4_2_clean=['rm -f %s.paired.bam'%sample]
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)]
post_5_clean=['rm -f %s.paired.sorted.bam'%sample]
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)]
post_6_1_clean=[]
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)]
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]
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)]
post_7_clean=['rm -f %s.withRG.GATKRecalibrated.bam'%sample,'rm -f %s.withRG.GATKRecalibrated.bai'%sample,'rm -f %s.Duplicates_metrics.txt'%sample]
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)]
post_8_clean=[]

cmdset=[]
for item in globals().keys():
    if item.endswith('_cmd'):
        cmdset.append(item)
cmdset.sort()

#########################################################################################
#write PBS file

###############Entry Point
try:
    cmd_entry=cmdset.index('step_'+step+'_cmd')    ##entry point
except ValueError:
    sys.exit('ERROR: STEP not recognized')

def _parsecmd(cmd):
    '''pase cmd str for step information'''
    info=cmd.split('_')
    cstep='_'.join(info[1:-1])
    return cstep

###########################
##headers
outfile=open(pbspath,'w')
outfile.write('#! /bin/sh\n')
outfile.write('#PBS -V\n')
outfile.write('#PBS -N %s\n'%tag)
outfile.write('#PBS -j oe\n')
outfile.write('#PBS -o %s\n'%logpath)
for item in refdict['--PBS--']:
    outfile.write('#PBS %s %s\n'%(item,refdict['--PBS--'][item]))
outfile.write('#PBS -d %s\n'%outpath)

#commands
outfile.write('echo "Job start: `date`"\n')

for i in range(cmd_entry,len(cmdset)):
    cmd=cmdset[i]
    cstep=_parsecmd(cmd)
    clean_flag=1
    outfile.write('echo "step %s start: `date`"\n'%cstep)
    outfile.write('if\n')
    outfile.write('\t%s\n'%('\n'.join(eval(cmd))))
    outfile.write('then\n')
    outfile.write('\techo "step %s done: `date`"\n'%cstep)
    outfile.write('else\n')
    outfile.write('\techo "step %s ERROR"\n'%cstep)
    outfile.write('\texit\n')
    outfile.write('fi\n')

    if keepmed_flag=='yes': #if so, keep files by force
        clean_flag=0
    
    if clean_flag==1:
        outfile.write('\n'.join(eval('post_%s_clean'%cstep)))
        outfile.write('\n')

outfile.write('echo "PIPELINE FINISHED"\n')
outfile.close()
######################################################################

if submit in ['False','FALSE','false','F','NO','No','no','n']:
    jid='Not_Submitted'
    logpath='None'
elif submit in ['True','TRUE','true','T','YES','Yes','yes','y']:
    cmdstr='qsub %s'%pbspath
    cmd=cmdstr.split()
    cmdout=subprocess.Popen(cmd,stdout=subprocess.PIPE,stderr=subprocess.STDOUT)
    jid=cmdout.stdout.read().strip()   ##JOB ID
else:
    sys.exit('ERROR: submit parameter not recognized')
 
print '#!#%s'%tag
print 'BAM\t%s'%bampath
print 'Entry\t%s'%step
print 'PBS\t%s'%pbspath
print 'JOB\t%s'%jid
print 'LOG\t%s'%logpath