TopHat options for small RNAseq

[Jeannine]

Hi,

We recently performed a small RNAseq for the first time at our institute (and NGS core) and unfortunately the quality of the reads is not ideal. However, it would be nice if I can still get something out of the data...
Therefore, I was wondering if I could change the tophat options used for alignment of the reads (at the moment the default options are used), i.e.if there are certain parameters specific for small RNA analysis?
Or is there anything else to consider? (after tophat I'm only counting the reads and compare these between the different samples, so tophat seems to be the crucial step for the analysis).

Thanks in advance,
Jeannine

[kcchan]

You probably don't need to be using TopHat for small RNA analysis since there is no splicing involved; Bowtie or BWA should work just fine. If you're concerned with low read quality, you may either filter out the low quality reads or ignore the quality information altogether using --ignore-quals in Bowtie. I would not recommend doing this, however, as the alignment tools use the quality data to determine the most accurate alignment. Ignoring the quality information may introduce some false positives during the alignment

[dvanic]

We use BWA for small RNA-seq, and an added advantage is that it will clip reads for you based on quality (i.e. dynamically). I would suggest trying it and seeing how many reads map.

Also, have you remembered to clip the adaptors? (asking just in case, sorry if it's obvious to you).

[illuminaGA]

Sorry if this is a dumb question.

How should I choose the GFT file for small RNA seq analysis? Can I use the same gene annotation GFT for regular RNA-seq pipeline? Thank you in advance.