Trouble shooting a low quality RNA-seq run/modofications

[exo]

Dear All,

I've recently started performing RNA-seq experiments to sequence whole viral genomes. I started with using the Script-Seq kit (originally sold by epicenter). I start the protocol with total RNA (required according to manual: mRNA or ribosomal depleted RNA) but limited to the input concentration.
I follow the protocol as suggested with a slight modification. I use 0.85X AmpureXP beads after PCR enrichment instead of 1X.

The quality of the run was quite bad. Clustering was on the upper limit for the V3 chemistry (range 1200-1400 k/mm2) with 86% passing filter. I was surprised to see the Qscore! Is it because there is a big portion of the fragments shorter than the read length?

Has anyone modified this protocol in such a way to make it more feasible on a 300x2 bp run? Is there any other step in the protocol that might lead to those short fragments?

I would appreciate the help with this! I attached an agilent trace for the libraries used (#1,2,11 did not work) and some screenshots from the analysis software.

Thank you!
Exo

[nucacidhunter]

I follow the protocol as suggested with a slight modification. I use 0.85X AmpureXP beads after PCR enrichment instead of 1X.

Using less bead would only increase the size cut off with no impact on library quality.

The quality of the run was quite bad. Clustering was on the upper limit for the V3 chemistry (range 1200-1400 k/mm2) with 86% passing filter. I was surprised to see the Qscore! Is it because there is a big portion of the fragments shorter than the read length?

High density clustering and sequencing through adapters will reduces overall quality scores, the insert sequences quality looks good. Quality should be checked after removing adapters from reads and there should be some improvements in quality scores.

Has anyone modified this protocol in such a way to make it more feasible on a 300x2 bp run? Is there any other step in the protocol that might lead to those short fragments?

This method uses random priming and average insert size is 200bp. To increase insert size one can try reducing fragmentation time to increase fragment size and using dilution of cDNA primer. This would reduce primed cites resulting in larger inserts.

[exo]

Thank you for the reply!

Indeed tweaking the bead to sample ratio will only affect the final library size rather than the quality of the library itself.

I think the problem in those libraries was the preferential clustering of small fragments compared to the large ones in the libraries. This definitely affected the overall run quality - especially with a 300x2bp chemistry.

I will go with your suggestion of decreasing fragmentation time - thanks again for the input!