I have shallow whole genome sequence data that I used the tool iChorCNA to find the mean copy number segmentation per region. The tool produces a seg file and can be seen here: https://github.com/broadinstitute/ichorCNA
I have also run these same samples to higher coverage on a targeted panel and used the tool VisCap Cancer to find the mean copy number segmentation. The tool can be found here and also produces a seg file: https://github.com/pughlab/VisCapCancer
I am looking to compare the outputs of each tool to see how close they overlap with each other. The problem is that iChorCNA produces outputs in bins that span the genome per million bases, and VisCap Cancer is only looking at the sites for targetted probes. These probes are anywhere from 1.3 million basepairs to 260 million basepairs in length.
What I am looking to do is segment the VisCap output into bins that match the bins used by iChorCNA. I want to then take the sum of the mean value given by iChorCNA at each bin that overlaps with the VisCapCancer results, and divide it by the total number of bins that fall in the range of the results to find the average iChorCNA reading per VisCap reading. Then, I can do a t-test, ANOVA or other measure to see how closely these two numbers are to each other.
I was wondering if anyone has directly compared copy number results from two data sources such as this before, such as shallow whole genome sequencing against a targeted panel, and if they have any advice as to how to go about this most effectively - including if it would be better to do the comparison using R, python or another language, and best practices for setting up the comparison most effectively.
Thank you for your time