Transcriptome assembly

[Mchicken]

Hey guys,

we are about to perform a transcriptome assembly approach. Here we want to apply de novo assemblers like Trinity, Oases and Trans-ABySS as well as reference genome guided assemblers like Cufflinks and Genome-guided Trinity. The aim is to build a global transcriptome of all assemblers by making a consensus. Several workflows were discussed in this thread http://seqanswers.com/forums/showthread.php?t=33452

We are about to perform paired-end sequencing but are not sure about the sequencing platform. We were offered two alternatives. The first one is the sequencing on a HiSeq2500, which produces 2x100 reads with an amount of 100million up to 180million read pairs. The second alternative is to sequence on a MiSeq, which would produce 2x250 reads with an amount of about 10million read pairs.

My question is which alternative you guys would prefer for a transcriptome assembly approach? Personally, I would prefer the first alternative for getting a high coverage, which is for example a filter criterion in Oases for the de Bruijn graph simplification and is also helping to capture more low abundant transcripts.

I`m looking forward to your recommendations.

Best regards
Mchicken

[sphil]

When you are using a HiSeq2500, can't you also go with 2x150 (Rapid mode?). Either way, imho it depends on for what kind of organism you are trying to identify the transcriptome. Rather complex one you might have to have some more reads so go with the HiSeq2500 option. If not, MiSeq should be just fine...

[Mchicken]

The assembly will be performed for the solanum lycopersicum (tamto) transcriptome which is rather complex. The rapid mode for the HiSeq2500 is unfortunately not offered by the sequencing institute.

So i think the 2x100 on the HiSeq2500 is the best alternative.