How to compare coverage eveness of two whole genome sequencing runs of the same sample?

Question

[xuang]

We are comparing two library prep kits: Nextera DNA Flex Library Prep and TruSeq DNA PCR-Free Library Prep for whole genome sequencing. The purpose of this comparison is to predict how will they affect the result of CNV calling.

So we whole-genome-sequenced the same sample prepped with either of the two preps, mapped them to the reference genome with bwa mem.

Now we have the general mapping statistics but that is not enough. To evaluate how CNV calling will be affected, our fist idea is to evaluate the eveness of coverage, but don't know to evaluate that properly.

Any ideas or existing tools to compare the coverage?

Answer 1

[Devon Ryan]

One simple method would be to run multiBamSummary (or make normalized bigWig files with bamCoverage and then multiBigwigSummary), after which you can use tools like plotCorrelation and plotPCA to see how well the (possibly normalized) coverage compares across bins in your dataset.

[rrdavis]

In this case which normalization method would you recommend for WGS

[Devon Ryan]

RPGC (aka 1X) normalization would be the simplest.

Answer 2

[Carlo Yague]

There is a rather simple tool designed to do exactly what you want: cnv-seq. Its a bit old but rather straightforward. I used it in the past and as far as I remember, it was easy to use and it made nice figures.