Low alignment percentage

[herber4]

Hi Weeks lab,

I am trying to snapshot whole transcriptome dms-mapseq data by mapping to a highly expressed transcript in our controls. We are getting extremely low - to almost 0% - alignment using shapemapper2. Do you have any idea as to what's going on?

Our data is paired-end, nexterra prepped from the whole transcriptome with a dms treatment of ~1%. Do I need to align outside of shape mapper and use the shapemapper_txome ?

Best,

Austin

[Psirving]

I have never used shapemapper_txome, but I can help you troubleshoot ShapeMapper2. However, I'll need more detail. What is the goal here, what do you mean by "snapshot"? Are your libraries ribosome-depleted or mRNA-enriched in any way? What is "almost 0%" exactly? How many reads align, and how many reads were collected? What percentage alignment were you expecting?

A few initial thoughts:

• You could try rRNA to confirm data quality. Even if depleted, rRNA sequences are usually the most abundant.
• 1% is nearly 0% by some standards, but would be a decently high percentage in a transcriptome experiment for a single transcript.
• Is there any discrepancy between modified and unmodified reads aligning? DMS can potentially introduce a lot of bias in transcript-to-transcript yields. The unmodified sample might be closer to RNA-seq abundances, but DMS may deviate substantially.
• Nextera preps discard many sequences at the 3' and 5' end (50 nt or so), so if your target is short, it may be underrepresented.