I have run Tophat on several Zea mays fq files and realize that there is a small percentage of rRNA contaminants within them. I know I could have removed them earlier by using ARTseq contamination protocol. However that is before Tophat. Now that I have run Tophat is there a way to remove the contaminants? I also plan to run cufflinks and cuffdiff. Do either of those have a way to remove the contaminants? Thanks!
Dear Pricilla,
You can get rid of contaminant reads from BAM file by using NGSutils program called bamutils: http://ngsutils.org/modules/bamutils/filter/
But you need to provide a text file with the names of those contaminant read and use option -blacklist fname
I hope this helps a little.
Shab
http://ngsutils.org/modules/bamutils/filter/
Alternatively, you can map your fastq file first to the rRNA using bowtie2 *without* the --no-unal flag, remove all the reads that align perfectly to the rRNA from the sam file, and convert it back to fastq. This actually saves you time as well.
There is a option in cufflinks " -M (mask file) ". You can provide gtf file of all non-coding RNA, which will be masked in transcript assembly.