Problems with Nextera

[John.Sawyer]

As a core we've performed several tests with epicentre's Nextera DNA Sample Prep and compared to our current Illumina preps. We are noticing very rigid coverage lines and significant bias in our reads with Nextera. I was looking forward to the short prep but the bias is not acceptable. I'm guessing it's the enzymatic cutting that's causing this. Enzymes have some design behind them that physical fragmentation doesn't have. Just wanted to put this out there for people not aware of the bias. Comments from other people's experience would be great.

[epibio]

The data we have seen indicate Nextera libraries are comparable to those made using other mechanical/enzymatic fragmentation systems. This includes full-coverage libraries from human, bacteria, Drosophila, soy, viruses, PCR amplicons, cDNA, and a variety of plasmid DNAs. Some of our collaborators have submitted a manuscript directly comparing Nextera libraries to those made by mechanical and enzymatic fragmentation methods, wherein they characterized insertion bias, library complexity, genome coverage distribution, GC bias, and other library metrics. These data will also be presented at the CHI XGen Congress in March 2011.

In general, all systems have bias resulting from many contributing factors. For example, standard mechanical fragmentation library preps have bias resulting from the ligation step. Downstream emPCR or bPCR steps contribute to GC coverage bias as well.

Can you provide any specific information regarding the bias you are observing? This would be very helpful in troubleshooting.

[Boonie]

Nextera bias

If the DNA template is limited in quantity, could the bias in the Nextera output in some libraries be due to PCR amplification, even with only the recommended 9 cycles of amplification?

[epibio]

Yes, PCR can be a significant contributor to bias (in all methods of library prep). See Adey et al. doi:10.1186/gb-2010-11-12-r119