View Single Post
Old 04-29-2015, 07:47 AM   #2
David_H
Junior Member
 
Location: Italy

Join Date: Jul 2009
Posts: 3
Default

Hi Brian,

This looks great ... it might be very useful for me.

A couple of questions/comments,

First, I am, for my sins, trying to do de novo assembly with data generated by the NextSeq V1 chemistry. I have initial (fairly poor) assemblies made using SPAdes after adapter trimming, overlap of PE reads, quality trimming and error correction (all done with your tools). Is it a good idea to use such an assembly to build the recalibration matrices for the raw reads (and then go back and start the whole assembly process again)?

As an alternative, I was wondering whether the error/QS profiles for the V1 chemistry are sufficiently uniform that a general model could be used (for example the matrices that you made on your data). I suspect not as my initial QS by position distros were quite different from yours (less good in general and with a "bell shaped curve" where the median QS was highest around position 50), but I'd be interested to hear your opinion.

Anyway, I'll go and play with this new toy and let you know how it goes.

best

David
David_H is offline   Reply With Quote