coding exons/ repeats/ rRNA, tRNA, snRNA, snoRNA

[EADIE]

hello,
i have some small RNA reads data of Arabidopsis thaliana. i have mapped them to Arabidopsis thaliana genome. now i want to exclude these reads(coding exons/ repeats/ rRNA, tRNA, snRNA, snoRNA, etc) from further analysis. but the question is how can i know which reads are (coding exons/ repeats/ rRNA, tRNA, snRNA, snoRNA)?

[mrawlins]

You could filter based on the location in the genome that the read maps to. Using samtools you could write an application (Java has the best samtools bindings, perl is okay, C is harder) to read in the mapped reads, produce a list of reads that map to an excluded location, and output to a new file all the reads that map somewhere else. That's how I would approach it. There may already be some tools that do similar things, but I haven't seen any.

[malachig]

This seems like the kind of task that could be done with BEDTools.
Links: Download and Manuscript.

The manual describes a few functions in that might be handy. For example, 'IntersectBed' using the 'Reporting the absence of any overlapping features' option might do the trick. Similarly the 'SubtractBed' function might be useful.

I assume you will need to convert your mapped read files to BED format and get an annotation BED file containing the regions you wish to exclude. BEDTools seems to have functions to facilitate some file conversions.

[steven]

Right, i would do as mrawlins proposed, only working on bed files (smaller than sam).
Of course this implies that you know the genomic location of your tRNA/rRNA/etc.