Normalization of RNA-seq bias

[buthercup_ch]

Hi everyone.
I'm have a question…, very basic I'm afraid.
I just received the results of my RNA-seq sequencing reaction, mate-pair, and it seems that there was a problem during read 2 sequencing. They propose me to start analyzing reads 1 while waiting for the repetition of the repeat the run.
I don't really like this procedure. I would be introducing a huge bias right? Is there any normalization able to get rid of such a bias?
Thank you in advance for your comments.

Cristina

[dpryan]

Assuming you're interested in doing standard differential expression, you can either introduce a batch factor in your model or use something like SVA. The general trick is to not try to directly get rid of the bias, but rather to include some metric of it in your statistical design.