In robinjugas/CNproScan: CNproScan detects and annotates CNVs in bacterial genomes.

CNproScan Vignette

CNproScan is R package developed for CNV detection in bacterial genomes. It employs Generalized Extreme Studentized Deviate test for outliers to detect CNVs in read-depth data with discordant reads detection to annotate the CNVs. It was tested and proven to be able to detect short CNVs. Following text is a workflow showcase. The CNproScan consist of a single function CNproScanCNV which carries the whole procedure. The steps necessary to get input files is also explained at the GitHub repository.

1. Website and issues

CNproScan github repository and issues reporting: https://github.com/robinjugas/CNproScan

2. CNproScan workflow

2.1 Install CNproScan

Install CNproScan from Bioconductor:

## try http:// if https:// URLs are not supported
source("https://bioconductor.org/biocLite.R")
biocLite("CNproScan")

or install the most current release from Github:

install.packages("devtools")
library(devtools)
install_github("robinjugas/CNproScan")

2.2 Reads alignment and coverage calculation

Apply the following steps to get the files needed to CNV detection.

# Prerequest: reference fasta file, samtools, bwa aligner
# Alignment
bwa index -a is reference.fasta
samtools faidx reference.fasta
bwa mem reference.fasta read1.fq read2.fq > file.sam
# BAM processing
samtools view -b -F 4 file.sam > file.bam # mapped reads only
samtools sort -o file.bam file1.bam
samtools index file.bam
# Calculate coverage with zero coverage reported
samtools depth -a file.bam > file.coverage
# Genome mappability file by GENMAP (https://github.com/cpockrandt/genmap) 
# only for mappability normalization
genmap index -F reference.fasta -I mapp_index
genmap map -K 30 -E 2 -I mapp_index -O mapp_genmap -t -w -bg

2.3 CNV detection

library("CNproScan")
# Working directory with files
setwd("workdir")
# File paths
fasta_file <- "reference.fasta"
bam_file <- "file.bam"
coverage_file <- "file.coverage"
bedgraph_file <- "mapp_genmap.bedgraph"

# For only GC normalization
DF <- CNproScanCNV(coverage_file, bam_file, fasta_file, 
                   GCnorm=TRUE, MAPnorm=FALSE, cores=4)
# Without any normalization
DF <- CNproScanCNV(coverage_file, bam_file, fasta_file, 
                   GCnorm=FALSE, MAPnorm=FALSE, cores=4)
# Both GC normalization and mappability normalization
DF <- CNproScanCNV(coverage_file, bam_file, fasta_file, 
                   GCnorm=TRUE, MAPnorm=TRUE, bedgraph_file, cores=4)