Drop down in Q30 on Miseq

[Elham1]

Hi All,

I run 2X 300 cycles Miseq kit. The Q30 score was drop down after 250 cycles in first read to 70% and remind drop down in R2 read to 50%. The deep coverage was also drop down. I am shocked to see dropping the coverage.

Any suggestion?

[GenoMax]

It is not so shocking. What kind of a run was this? What was the cluster density and type of reagents (V2/V3)?

One of the prime reasons of Q-score drops are generally overloading, low nucleotide diversity and/or short(er) inserts than what you expect.

Take a look at these threads:

http://seqanswers.com/forums/showthread.php?t=39760

http://seqanswers.com/forums/showthread.php?t=40976

Illumina's published MiSeq technical specifications are roughly in the ballpark: http://res.illumina.com/documents/pr...heet_miseq.pdf

[Elham1]

I used V3 kit. The cluster density is 792 (K/mm2). I didn't use the low nucleotide diversity.

[GenoMax]

Take a look at the data using FastQC. Post the results. For V3 kits, the cluster density is fine. You could have short(er) inserts than you expected.

Brian has a way to plot insert size distributions with BBMap tools. Search the forum for the exact command line to use.

[GA-J]

Hi All,

I run 2X 300 cycles Miseq kit. The Q30 score was drop down after 250 cycles in first read to 70% and remind drop down in R2 read to 50%. The deep coverage was also drop down. I am shocked to see dropping the coverage.

Any suggestion?

Hi, Elham1, what happened on your "Drop down in Q30 on Miseq" run? Did you find a way to solve it? I just run in to the same situation. Would like your suggestion.Thanks.

My run is 2X300 cycles of Miseq V3 kit with 16s rRNA seq amplicons. The cluster density is 1079k/mm^2. The Q30 score drop down from 80% to 70% at 250 cycles in first read and still dropping down in R2 ?%(still running).

Welcome any comments.

[GenoMax]

My run is 2X300 cycles of Miseq V3 kit with 16s rRNA seq amplicons. The cluster density is 1079k/mm^2. The Q30 score drop down from 80% to 70% at 250 cycles in first read and still dropping down in R2 ?%(still running).

Welcome any comments.

Take a look at this thread: http://seqanswers.com/forums/showthread.php?t=59558. Contact Illumina tech support once the run completes.

[GA-J]

GenoMax,thank you very much!

We called illumina today, suggested to decrease the library and increase the Phix for bringing down the cluster density to ~800k/mm^2. A replacement of cartridge will be sent.

Still not sure what happened. After reviewed the data with Miseq reporter, unclassified rate is very high, 70-80%. Will search for more information to find answer.

[microgirl123]

Did you classify down to species or just to genus? I've only once tried to classify down to species once on the MiSeq, and it returned most of the samples as unclassified (even at the genus level). When I reran the analysis only to the genus level, it worked fine.

[GA-J]

Sounds reasonable! My data on species is 0. But how do I set analysis only to genus level? I am new on data analysis. I will review miseq reporter user guide.Thank you!

[microgirl123]

Hmm... I seem to remember there being an option to click species-level classification in IEM, but now it looks like it's genus-level.

Truthfully, most of our users do their data analysis off the MiSeq with other software such as MOTHUR. The documentation for what MiSeq Reporter is actually doing when it's classifying your samples is sparse, and I'm not sure what, if any, quality control is done on the data before classification. There's a MOTHUR SOP online for MiSeq data, which is a pretty good starting point since it walks you through the entire process.

[GA-J]

Fantastic! I will work on it, thank you again, Microgirl123 !