Solexa GII transcriptomic data

[johnnyh]

Dear NGS community.
Please can you give advice on the following:

I have 24Million paired end 50bp reads from human GII transcriptomic sequencing. I would like to know the following:

1) Assemble reads into full length transcripts

2) Identify which transcripts are novel (ie, without having to manually view the whole human genome in a browser).

I have read your pages on software and see programs like velvet and tophat that could do step 1. Does anyone have experience of doing this type of analysis and know a basic pipeline I could follow?

Thanks in advance of any advice.

Kind regards,

John.

[Torst]

John

I have 24Million paired end 50bp reads from human GII transcriptomic sequencing. I would like to know the following:
1) Assemble reads into full length transcripts
2) Identify which transcripts are novel (ie, without having to manually view the whole human genome in a browser).

You have 24*50 = 1200 Mbp of sequence. The human transcriptome is about 100 Mbp. Therefore you have 12x average coverage. The paired end insert size is 200 bp. Given the number of splice variants and repeats greater than 200bp it will be a challenge to achieve your goals easily. Most of the transcripts you identify will be PARTIAL and/or SHORT.

[johnnyh]

Hi Torst

Thanks for answering my question.
I will let you know how I get on, I will first do a stringent assembly with EDENA (Hernandez et al. 2008) and see what that looks like. Even if these are short and partial transcripts, it will be useful to know which of those are novel.

I wonder if anyone ever repeat masks the reads first, but I guess a lot of sequences will just disappear doing that.

I will brain storm to find how best to quickly do identify novel transcripts.

Thanks again for taking the time to answer.

Kind regards,

JohnnyH

[baohua100]

Mapping to the genome and to see how many reads mapped to the exons, introns and intergenic region.

You can then use reads that mapped to the intergenic region to identify novel transcript.

[johnnyh]

Hello Baohua100,
Thanks for your message,
How to do that in a pipeline?
I used maq and bowtie to map to the genome but to quickly find which mappings are in introns/intergenic regions, how to do that?

Thanks.

johnnyh