DeCovA requires at least R and bedtools/GATK softwares to be installed; additionnally, it can use picard-tools (for deduplication), samtools (for mapq filter), and GATK (alternatively to bedtools; required if a base-q filter is needed; GATK is also aware of pair reads overlap). The script will first attempt to run programs installed as root with the following names: samtools, bedtools, picard-tools, GenomeAnalysisTK; if not found, it will try to find them according to the paths provided in the command-line.
DeCovA also requires perl modules: IO::Compress::Gzip.
An annotation file needs to be provided (-r option), for all the options that use gene coordinates: UCSC refgene.txt or Ensembl .gtf/gff files are OK.
ex:
wget http://hgdownload.soe.ucsc.edu/goldenPath/hg19/database/refGene.txt.gzwget ftp://ftp.ensembl.org/pub/grch37/release-92/gff3/homo_sapiens/Homo_sapiens.GRCh37.87.gff3.gzwget ftp://ftp.ensembl.org/pub/grch37/release-92/gtf/homo_sapiens/Homo_sapiens.GRCh37.87.gtf.gzDeCovA can be executed via command-line execution of the main perl script:
perl path/to/DeCovA/bin/DeCovA [options]the script can be changed to executable:
chmod 755 path/to/DeCovA/bin/DeCovAthen the command is :
./path/to/DeCovA/bin/DeCovA [options]The DeCovA script directory can be added to the $PATH :
echo 'export PATH=$PATH:/home/me/path/to/DeCovA/bin/DeCovA' >> /home/me/.bashrcthen the command is only :
DeCovA [options]DeCovA can also be installed:
cd path/to/DeCovA/
perl Makefile.PL
makethen, as root:
sudo make installthen just enter
DeCovA [options] -f / --file [file]: list of bam files (comma separated, or set several times)
-F / --fList [file]: file with such a list of bam files (one bam per line)
-d / --dir [dir]: directory(ies) where to find bam files (comma separated, or set several times)
-s / --suffix [str]: suffix to add before opening bam files
-r / --ref [file]: gene annotation file (can be .gz)
--fmt [gtf/gff3/ucsc] : gene annotation file format (ucsc <=> UCSC refGene) ; if not provided, determined from extension (txt => UCSC refGene)
-b / --bed [file]: bed file, used to analyse depth coverage
-m / --mut [file]: mut file, used to plot known mutations ; format: "chrpos(1-based)info" (vcf files are ok ; can be .gz)
-i / --id [str]: list of of genes/transcripts ids (comma separated, or set several times)
-I / --idList [file]: file with a list of of genes/transcripts ids (one id per line)
-g / --genome [file]: path to genome.fa file, if available (required if using GATK)
--sex_file [file]: format: patientsex
--raw_cov [file]: use this coverage tool output .cov file (to skip bam analysis)
--bed_cov [file]: use this DeCovA's output .cov.txt file (to skip cov bed analysis in CNV detect)
-O / --outdir [dir]: out directory (default: folder named with date)
-S / --graphSum : will perform graphSums (sum of covered samples by position)
-A / --allSample : will perform graphAllSample (depthline by gene and by sample, all samples graph on same .png file)
-X / --bySample : will perform graphBySample (depthline by gene and by sample, one sample by .png file)
-M / --noDepthMut : does not print, foreach file, depth at known mutations provided by opt -m (default: yes if -opt m)
-P / --covPlot : will perform covPlots
-B / --covBed : will output cov of bed intervals
-C / --CNV : will output CNV foreach bed intervals
--Reseq : [float,0-1] : print bed interval if cov < value (def: do not print)
--geneReport : will print all uncovered genomic intervals (within gene region) in 1 txt file per sample (default: no)
--bedReport : will print all uncovered intervals (within bed intervals) in 1 txt file per sample (default: no)
--summary [Y/N] : to print summary txt file (default: yes if -S -A -X)
-k / --keepCov : do not erase coverage file at the end of the process
-K / --keepBed : do not erase bed file inferred from gene list, at the end of the process (and eventually rename)
*gene/transcript regions analysis param.:
-N / --nonCoding : analyse also Non coding transcripts (default: no)
-U / --noUTR : does not take into account UTR regions, for graphs (default: yes)
-u / --noUTRinTxt : does not take into account UTR regions, for summary txt file and plots (default: yes)
-t / --depthThreshold [int]: depth thresholds (comma separated, or set several times)
-T / --printThreshold [int]: depth threshold used for txt outputs (must be one of those in opt -t; default : the smallest one)
--noGraphThreshold : all graphs will be printed, whatever the coverage (default: only the genes not fully covered at threshold in -opt -T will be drawn)
--noAllTranscripts : does not print All transcripts on same file, in graphBySample (default: yes)
--maxDepth [int]: max depth value when printing graph (optional)
-l / --expand2val [int]: length to add at each ends of exons, on graphs (default: 0) ; or [int1,int2] : lengths to add in 5' and 3'
--UDstream [int]: length to add at each ends of genes, on graphs ; or [int1,int2] : lengths to add upstram and downstream
--splitBedFromId : if padding creates overlapping exons, take the mid between them (for report)
--mergeBedFromId : merge overlapping exons
-L / --expand2bed : expand length of gene analysed regions to bed coord, if -l < bed , on graphs (default: no)
--Ltxt [+/-int]: does take into account expanded length (from -l and -L) for txt outputs (default: no), or add a different length
--UDtxt [+/-int]: does take into account up/downStream length for txt outputs (default: no), or add a different length
-R / --noReverse : does not reverse regions if sens of transcript = (-) (default: yes)
--nGraph : max nber of graphs per sheet (default : all samples or all transcripts)
*plot param:
--binPlot [int]: bin width for covPlot (default=10)
--maxPlot [int]: max depth for covPlot (default=100)
--genePlot : will perform plots for regions extracted from genes coord, not only for bed intervals (default: no)
--interPlot : will produce intersection covPlot (default: no)
*bam filters
--dedup : do not take in account dup reads (default keep all reads; enter "do" to perform Picard deduplication)
--mbq : minimum base quality (default 0; requires gatk)
--mmq : minimum mapping quality (default 0)
*cov_bed param:
--cov_fields [min/max/tot/mean/median/cov]: fields foreach intervals in covBed (comma separated) (default: min,mean,cov)
--Lbed [int]: length added out of bed interval ends (default: 0)
--split_bed : splits overlapping bed intervals for Cov and CNV analyses
--no_overlap_bed : removes overlapping bed intervals for Cov and CNV analyses
--cut_bed [+/-cutL:x,minL:y,maxL:z,keepLast:s]: cut bed intervals in shorter fragments:
cutL : length of segmentation (def: 150)
minL : min length required to keep the last interval, after segmentation (def: --cutL/2)
maxL : length above which bed intervals will be segmented, in N segments of "cutL" length (def: as --cutL)
keepLast : if last interval shorter than minL :
enter m (merge) if want that last two ones are simply merged
enter h (half) if want that last two ones are output with length = half of their sum
enter n if want to through it out
--reAnnot_bed : removes and replaces 4th column of bed file with gene info (optional args: g,t,e,i,o : indicates to annotate with gene/transcript/exon/intron/intergenic infos; default: all)
CNV_detect param:
--level2 : "avg"/"med" : use average/median as center of depths of a region (def: med)(if spread2 is set, level2 is unset, unless explicitedly)
--spread2 : "std"/"qtile" : use standard deviation/deviation from quartile as dispersion of depths of a region (def: none)(std forces avg, qtile forces med)
--level_del [float [0-1]] (def: 0.8)
--level_dup [float >1] (def: 1.2)
--spread_del [float <0] (def: none)
--spread_dup [float >0] (def: none)
--range [float]: samples kept for avg-std calculation if within mediane+/-range*quartile (def: none, ie all samples used)
--highQual [li:float/ls:float/si:float/ss:float/c:int]: flag as high qual if one of following criteria, comma separated : li=level inf, ls=level sup, si=spread inf, ss=spread sup, c=consecutive ; ex : li:0.25,ls:1.75,si:-5,ss:5,c:2
--ex_region [float [0-1]] : region excluded from analysis if CNVs/N_samples >value (def: 1)
--ex_sample [float [0-1]] : sample excluded from analysis if CNVs/N_regions >value (def: 1)
--ex_cov [float [0-1]] : region excluded from analysis if none of the samples have cov >=value (def: 0)
--ex_DP [int] : region excluded from analysis if avg depth <=value (def: 0)
--max_nonCNVcons [int]: max nber of non-CNV consecutive intervals tolerated within a CNV (def: 0)
--max_nonCNVrate [int]: max rate of non-CNV intervals tolerated within a CNV (def: 0)
--ratioByGender [a/g/no]: enter "a" : foreach region from all chrom, depth ratio computed separately for F and M ; enter "g" : foreach region from gonosomes only, depth ratio computed separately for F and M. def: no (depth ratio for F and M together)
--normAllChr : total depth used to norm sample depths = sum on all chr, whatever the sex (def: double the depth for chrX if male, and skip chrY in the sum)
--normDepth [mean/tot] : total depth used to norm sample depths = sum of total depths of each region or sum of mean depths of each region (def)
--graph_byGene : to enable graph for gene affected by a CNV (def: no)
--graph_byChr : to enable graph by chromosome (def: no)
--graph_byCNV : to enable graph around each CNV (def: yes)
--CNV_fields [min/max/med/avg/std/Q1/Q3]: list of fields foreach region (comma separated) (default: none)
*external tools path:
--bedtools [dir/file]: enter path to executable, if not installed as root or not in path
--samtools [dir/file]: enter path to executable, if not installed as root
--picard [dir/file]: enter path to executable .jar, if not installed as root
--gatk [+/-dir/file]: cov analysis will be performed by gatk (default:bedtools; enter path to executable, if not installed as root)
*general:
-x / --ram [int]: memory for gatk (in Go)
--cpu [int]: multi-thread for gatk (def: 1)
-v / --version : current version
-h / --help : help
./path/to/DeCovA -d path/to/bam_dir/ -r path/to/Refseq.txt -b path/to/targets.bed -M path/to/mut.list -t 20,50,100 -A -S -P -B -C./path/to/DeCovA -f path/to/file1.bam -f path/to/file2.bam -r path/to/Refseq.txt -b path/to/targets.bed -M path/to/mut.list -t 20,50,100 -A -S -P -B -C./path/to/DeCovA -f path/to/file.list -r path/to/Refseq.txt -b path/to/targets.bed -M path/to/mut.list -t 20,50,100 -A -S -P -B -C./path/to/DeCovA -d path/to/bam_dir/ -r path/to/Refseq.txt -i GENE1,GENE2,NM_xxx1,NM_xxx2 -M path/to/mut.list -t 20,50,100 -A -S -P -B -C./path/to/DeCovA -d path/to/bam_dir/ -r path/to/Refseq.txt -i genes.list -b path/to/targets.bed -t 20,50,100 -A -S -P -B -C