What is the best pipeline for RNASeq atm?

[foxyg]

Hi guys,

What do you recommand for RNASeq analysis. I know some people use Tophat and cufflinks. Is tophat + cufflinks the best way to go?

Thanks

[GKM]

The answer to your question very much depends on what the goal of your analysis is.

And even if we knew it, in most cases, there would be no objective answer

[mrawlins]

If you are working in eukaryotes and are interested in splice variants, tophat + cufflinks is a good tool. I haven't been able to get the pipeline to work with SOLiD data, though.

For SOLiD data, the statistical methods outlined in Bullard et. al. (2010) work very nicely if you want to calculate differential expression. I don't know of any software package that does those, but you can program them in R or Perl without too much trouble.

If you have other aspirations for your data, you will likely have to create at least a few of your own tools to get an analysis that fits your experiment type. This field is still new enough that many types of analysis haven't been standardized yet.

[foxyg]

I am using human samples donig control and cancer group comparison

[GKM]

I am using human samples donig control and cancer group comparison

That's not a sufficient amount of information. Do you just want to look at expression levels for known genes, or you're interested in isoforms, unknown transcripts, chimeric transcripts (relevant for cancer), etc.

[foxyg]

I want to look at all the thing I can look at for the data give. BTW the data obtained from Illumina seqencer.

[GKM]

Then you are looking for a combination of de novo splice mapper and a transcript reconstruction and quantification program. TopHat + Cufflinks is one way to go but by no means the only thing out there