illumina de novo txome assm: overlapping paired ends or long inserts?

[Zigster]

There seems to be some debate about the optimal paired-end read length for de novo transcriptome assembly.

Some are advocating selecting for short fragments - so short you can see overlap between pair mates. Others contend there should be bigger inserts between the ends in order to build the scaffold and leverage pairing information.

I would like to hear your views on this.

[MattB]

Hi Zigster,

Interesting question you pose, we have had some paired-end transcriptome sequencing done and for some reason have quite a small average insert size (ie. lots of overlap between pair mates). It's hard to say what impact this has had on contig/scaffold building, since we haven't got anything to compare to! Like you, I would be interested to hear what others have observed, as it might influence our strategy for futher transcriptome sequencing that we plan to do.

Matt