Detecting Dispersed vs tandem duplicated

Question

[znl207]

I am working with CNVs in a population, I've used a couple of different detection methods, I am currently using lumpy and delly- I am wondering if folks have any recommendations for distinguishing between dispersed and tandem duplications from the output of various CNV/SV detection tools. How would you go about detecting if a duplication is tandem vs. dispersed? Thanks!

Answer 1

[piechota.marcin]

For tandem duplication you will find reverse order pairs around breaking points. Left read just before ending breaking point and right read just after starting breaking point. For dispersed duplication you will find larger than expected insert size for read pairs around breaking points. One read within borders of CNV and second read outside borders of CNV (pairs can be reverse or in order).