Sample pooling, do FFPE derived samples need to be separate from blood derived?

[Iikki]

Hi!

We sequence both blood/fresh frozen and FFPE derived exomes using the Illumina HiSeq2000 and HiSeq4000. After the sample prep the samples are pooled. We have always pooled FFPE derived samples separate from blood/FF derived, since FFPE samples tend to produce a bit less data and we have been worried that the amount of data from these samples would be further reduced when competing with better quality DNA. I have no idea why this would happen though. The molecules bind through their adapter sequences, which are the same and "fresh" regardless of tissue of origin, and the amount of input material on the flow cell is naturally the same. Does anyone have experience on running FFPE and blood/FF samples on the same lane or any idea why we should not do this?

BR, Iikki

[kerplunk412]

As you noted, unless you are doing PCR-free library prep, the samples you put onto the sequencer are all fresh, high-quality PCR products and thus the type of starting material shouldn't really matter. That being said, shorter library molecules do form clusters more efficiently, so that should be taken into account when mixing.

It makes sense to me that the FFPE samples would have lower mapping rates and more errors, so you may need to sequence them deeper to achieve the same data quality as the blood samples. However the actual clustering efficiency of the samples should not be any different, assuming PCR was used in the library preps.

[Iikki]

Thank you kerplunk412!