Skip to content

Latest commit

 

History

History
96 lines (65 loc) · 3.06 KB

CNV-Seq.md

File metadata and controls

96 lines (65 loc) · 3.06 KB
Raw

Tutorial to detect copy number variation using CNV-Seq

Table of Contents

About the tool

CNV-Seq allows the detection of copy number variation using NGS data.

Input

In order to reduce the noise around repeat regions in the genome, first filter reads with a mapping quality > 4:

samtools view -b -q 4 sample.bam > sample.qfilt.bam

Hits file

We only need to provide two "hits" files for CNV-Seq to work, one for the sample, and one for the reference.

These can be extracted as follows:

samtools view -F 4 sample.qfilt.bam | perl -lane 'print "$F[2]\t$F[3]"' > sample.qfilt.hits

This creates a two column, tab delimited file, with the second column giving the corresponding 1-based leftmost mapping position of a read.

2L	1
2L	1
2L	4
2L	4
2L	4
2L	4
2L	4
2L	4
2L	4

Next, to select only fully assembled chromosomes run filt.pl. This will output filtered files eg sample.qfilt.hits.filt:

Run CNV-Seq

Now run the main perl script:

cnv-seq.pl --ref ref.qfilt.hits --test sample.qfilt.hits --genome-size 23542271 window-size 10000

Output

This produces two files sample-vs-reference.cnv and sample-vs-reference.count

sample-vs-reference.count shows the raw count data for each CNV.

chromosome start end test ref
X 1 363 70 124
X 183 545 82 123
X 365 727 90 115

sample-vs-reference.cnv contains the stats.

chromosome start end test ref position log2 p.value cnv cnv.size cnv.log2 cnv.p.value
X 1 363 70 124 182 -0.479319752689881 0.0462467378993667 0 NA NA NA
X 183 545 82 123 364 -0.239368959969129 0.194296291525077 0 NA NA NA
X 365 727 90 115 546 -0.00804341386267303 0.488268352520283 0 NA NA NA

Plotting

To plot and output CNV info for all samples, run cnv_seq_process.sh on .cnv files:

$ bash cnv_seq_process.sh *.cnv

This calls a modified version of the main script provided with CNV-Seq.

The main tweaks:

  • Change the plotting colours/densities
  • Show genes of interest in a closeup plot of notch region e.g.
  • Save CNV details for each file parsed for later processing

This will by default produce two plots (one for notch region, and one for chromosome X) and a cnvs.txt file. To plot for different chromosomes alter the chrom var in the script cnv_seq_process.sh.

Visualasing CNVs in IGV

To create an IGV compatible track for each sample, run cnv2gff.pl:

for file in *_cnvs.txt; do
	 perl cnv2gff.pl $file
done