Gene panel exome data

[yl01]

Thank you nexgengirl and Bukowski!

[nexgengirl]

See Bukowski answer. It came up as I was entering the previous post.

[nexgengirl]

I have not personally used the Haloplex platform before. I've analyzed a lot of exome data but not Haloplex.

From briefly looking at the information on the Agilent site it seems that first the DNA is fragmented with restriction enzymes which will be exact cuts (as opposed to "random" shearing in typical exome experiments), second the probes are hybridized (the way its described this also seems very specific since its to the ends of restriction fragments), third the targets are purified and ligated, fourth there is a PCR step, and then fifth sequencing.

From my interpretation of this setup I would think it would be okay to retain the PCR duplicates (since there's PCR in the 4th step to enrich for your targets) but I would hope someone else who has used the Haloplex technology could comment on this as well.

[Bukowski]

Thank you nexgengirl! I know it was using Agilent HaloPlex technology. So I think it is hybridization based. Right?

Sort of - there's also an amplification step. Read the protocol. You can't deduplicate Halo data, it's not like working with SureSelect. Effectively you get mini-amplicon regions that match up with the restriction digest sites used.

[yl01]

Thank you nexgengirl! I know it was using Agilent HaloPlex technology. So I think it is hybridization based. Right?

[nexgengirl]

How was the DNA prepared? Was it hybridization or pcr based design? If it was PCR-based you would expect high PCR duplication rate but if it was hybridization based then yes this is high.

[yl01]

Hello,

I have got exome sequencing data from 24 samples. It was done on a MiSeq sequencer. The data are not whole exome data, instead it is gene panel of about 200 genes. I have processed the whole data using the usual whole exome pipeline BWA+GATK, of course with the corresponding bed file. It turned out that the PCR duplication rate in majority of the samples is very high, 70% to 90%. This also caused the mean coverage to be very low, about 4. I am wondering if this is normal in gene panel data. And is it OK in gene panel sequencings to not remove PCR duplicates. Any other points regarding the processing of gene panel exome data is much appreciated.