Generation of BED/GTF/GFF-files

[StevenVanLaere]

Dear All,

I am relatively new to RNA-seq data analysis. I did my share of literature but theory and practice are two different things. So I encounter small problems/questions on which I cannot find straight forward answers. Here is one, hope you can help.

I have aligned my reads to the reference genome (hg19) with TopHat2 and now I want to use DeSeq(2) to identify differentially expressed genes. Obviously a genome annotation file (in this case GFF3) is needed. I wonder what the best solution is. Getting the annotation file from e.g. UCSC? Or generate one from the BAM-files I have, thus converting BAM to BED to GFF3 (to GTF)? In the latter case, how do I deal with the fact that I have multiple BAM-files, i.e. one per sample. I expect differential gene expression so I guess that a GFF3 generated from a BAM file of condition 1 will be different from those for condition 2.

Thanks for your help.

Steven

[GenoMax]

Where did you get your hg19 genome from? It would be best to get the GTF file from the same source.

One solution is to get GTF file from iGenomes. The downloads are large but the files (sequences, annotation, indexes are all in sync as far as the names etc goes) and this will save you time down the road by avoiding problems with annotations etc.