Transcriptome assembly to validate annotations

Question

[guillaume.rbt]

Hi everyone,

Several gene annotations have been made on the fungal species I'm working on.

I would like to cross and validate those data to get just one annotation, which would be the closest to the reality.

For this task I have several RNAseq sets of data, obtained in different conditions.

With that, I would like to perform transcriptome assembly, to get transcripts predictions, and confront them to the annotations.

I was wondering if it is a better idea to do transcriptome assembly on each RNA-seq independantly, and then gather the transcripts predicted, or to merge all the reads of all RNA-seq in one file, and to perform the assembly on this file.

Does anyone have an opinion on this?

Thanks

Guillaume

[st.ph.n]

You would be better off combining all of your data. You can map each sample back to the assembly to get abundance per sample/replicate/condition. If you want to do any kind of DGE analysis, this would be ideal.