How to use simultaneously long and short reads to assemble de novo a transcriptome?

Question

[friasoler]

Dear colleagues Thanks in advance for your help I would like to assemble, de novo, a transcriptome from which I have two kind of data:

1) long reads (Illumina Roche GS FLX System, single-end reads ~400bp) 2) Short reads illumina ~40bp.

I have used already the long reads for the assembly but the results are poor I have used the short reads (using Trinity) and the assemble outputs are much better but I would like to use both data at the same time hoping to build a much better transcriptome. I could not manage it using Trinity. Could you please give me some hints how to use both long and short reads to assemble the transcripts? Best regards Roberto

Answer 1

[Antonio R. Franco]

Obviously it depends on the program you use

Oases for example, uses categories or "channels" each one defined as long, short, paired, single, etc in the command lane. This let you use all kind of reads, and even a reference

If compiled by default, you can use 2 of these categories. But you decide the number of categories the program will be compiled. This can be done by editing the Makefile o following the instructions