High multiple mapping rate Drosophila pair-end 50 bp RNAseq

[TTYE]

Hi, I used Tophat2 for alignment. The mapping rate is ~92% while the multiple mapping rate is ~25%. Based on the FastQC result, it has some problem in the per tile quality, I see many red strips. So I trimmed the low quality head and tail, but the mapping rates and multiple mapping rates are not changed much. What would be the reason? How can I fix it? Thanks!!!

[GenoMax]

Post the FastQC plots (quality, adapter) to give us an idea of the "problem" you are referring to. Did you do trimming based on quality (what was the cutoff used) or just blanket trimming for a certain number of bases at beginning/end of read?

[TTYE]

per tile picture attached

Hi GenoMax, Thanks for your quick reply.

I should be more accurate, I trimmed the sequences based on the quality score. Do you think filter the whole sequence with low quality segment will be better?

Just looking at the per tile figure, what could be the reason of such situation. I checked the FastQC webpage, the author suggest that maybe the bubbles in the flow cell. What do you think?

[TTYE]

trimming parameters

I used trimmomatic with parameters:
LEADING:20 \
TRAILING:20 \
SLIDINGWINDOW:4:20

Thanks!

Post the FastQC plots (quality, adapter) to give us an idea of the "problem" you are referring to. Did you do trimming based on quality (what was the cutoff used) or just blanket trimming for a certain number of bases at beginning/end of read?

[GenoMax]

Have you asked the sequencing facility if there was any problem with the run this sample was on? What kind of sequencer (MiSeq/HiSeq) is this data from? Particular tiles have been affected across entire run so unless there were multiple bubbles stuck at certain positions in the flowcell (seems unlikely) ...

[TTYE]

Just confirmed by the sequencing core, their machine does go wrong around that time. So what should I do with it? Thanks!

[GenoMax]

They should not have released the data in the first place, if there was a known problem with the run. I suppose you can ask them to re-run your sample (if this was a machine/reagent problem then Illumina will generally provide free replacements, if your facility has a maintenance agreement).

Going back to the original question of multi-mapping .. that may be a characteristic of your sample. I would imagine that run related problems should not change the composition of your sequences (this can be validated if you do get a new run done from your facility).

[TTYE]

Got you. Thanks a lot!