Transcriptome analysis

[sn775]

Hi,

I am working with RNA-seq data. I assembled the illumina seq data in Mosaik using reference genes. The output was in the form of coverage (eg 10.5X, ....and so on for each gene). I want to proceed these data to find the differential expression. Previously I used DESeq (a package in R) to analyze the no. of reads. It works fine with those kind of data set but it is not compatible for my data now (because of decimal). Does anyone have any idea which software I can use to find the differential expression?
Thank you.

[westerman]

Not having used DESeq my answer may be totally off the wall but why not modify DESeq, either natively or via a pre-processing step to work with decimals? The easiest approach would be to just 'round' all numbers. Of course you would lose some precision with that method.

[chadn737]

Why do you want to use coverage rather than read counts for differential expression? Is it simply because of the type of output you get from Mosaik?

I am not familiar with Mosaik, but if you have a SAM/BAM alignment file then its easy to obtain the read counts for each gene using HTSeq-count which is then suitable for input into DESeq.

[sn775]

Thanks for your suggestion.
There is no problem in out put from assembly. I have both read data set and coverage data set. I wonder about the realiability of the dataset because of variation in gene size.

I will check my data set, by making them round.

Do you have any other suggestion for RNA-seq data - which software is better to assemble (with reference genes) and which statistical programme I can use?
Thank you once more.