exome cnv analysis of cases only

Question

[peris]

Hi all,

I have done exome sequencing off affected family members in 5 different families. I have done the family-wise SNV analysis on it. Now I want to do CNV analysis also on it. Since I don't have any control exome, can I draw some meaningful conclusion by doing family-wise CNV analysis only on the case samples.

Thanks and regards

[donfreed]

Is there a program or method you are particularly interested in using? My experience has been that read-depth based approaches do not work well with exome sequencing. Lumpy seems to work very well with exome sequencing, especially after alignment with BWA MEM.

[peris]

Hi @domfreed,

Thanks for the suggestion. I am planning to use conifer.

Answer 1

[Katie D'Aco]

Your job will be difficult without matched controls to weed out systematic false positives and normal/common CNVs. CNV calling from exome data is noisy, so the former will definitely be a problem. My suggestion is to download some reference CNV sets from dbVar to compare to your cases so you can remove common CNVs and identify any known pathogenic CNVs shared by affected family members.

Answer 2

[DG]

Do you have other exomes from other projects? You don't need controls from the family to do Conifer or some of the other CNV analysis programs tailored for exome data. You just need other exomes, generally captured using the same capture technology and sequenced on the same platform. The more the better (due to noise in the data). This helps normalize for both common CNVs as well as regions that just don't capture well.

[Dataman]

Hi Dan,

So far, I was under the impression that with CoNIFER you can only do the germline CNV analysis and NOT somatic CNV analysis. If I have understood correctly your answer, it is also possible to do somatic copy number variation analysis using ConNIFER by just feeding all the normal and tumor exome data to the tool. Is this correct? I am asking this since I am working on the Whole Exome Sequencing (WES) data where I have both blood and tumor samples and I was wondering what tool I could use to do somatic CNV analysis.

I have been using CoNIFER only on blood samples to find the germline CNVs.

Thanks in advance for you response.

Answer 3

[Vivek Todur]

Recently I have come across with a tool ExomeCopy, as per their manual, it can perform well on unpaired samples. Though I have not tried, but can be potential option to give try. ExomeCopy is basically R script and is located @ ExomeCopy

[Christian]

exomeCopy worked really well for me, but you need normal controls for filtering