Help with Repeat Variant Calls

[larangela]

Hello!

We conducted library prep and sequencing with Illumina. During data analysis, bioinformatics claims that the variants for all 40 samples were exactly the same. The %PF was >80, % aligned was about 3.64 (read 1) and 3.54 (read 4), with a density of 235. Illumina tech support stated that the run quality is passing but what would cause the same variants to be called for every sample in the run?

[Brian Bushnell]

Perhaps you can give more details, like the sequencing platform, experiment type, whether the samples are from the same individual, how multiplexing was performed, and so forth? Also, why you expect only 3.64% of the reads to align, which is incredibly low.

[larangela]

The TruSeq custom amplicon low input run (from PCR) was sequenced on a MiSeq for various individuals. We used Illumina's library prep kit as well. The align % is very low which was the only portion that seemed strange, per the Illumina Scientist. I'm not quite sure what could have caused this issue but it's occurred in a couple of runs.

[Brian Bushnell]

I would not use the data unless you can determine why the mapping rate was so low and ensure it would not cause incorrect results. Most likely, the samples were highly contaminated, which would also explain why all gave the same variant calls. I suggest you BLAST some of the 96% of nonhuman reads to public databases like nt and find out what they hit, and report back. Also, I'm glad it aroused your suspicion that all samples indicated identical variants; that does not usually occur outside of highly inbred villages, unless the variants are close together. It would be helpful if you could describe the methodology for variant-calling, and sample selection. Are the samples all from the same race, or same village? And how big is the amplicon region?

[nucacidhunter]

Hello!

The %PF was >80, % aligned was about 3.64 (read 1) and 3.54 (read 4), with a density of 235.

I wonder if those figures are from secondary analysis. Those to me looks like the sequencing metrics reported by Illumina SAV. %PF would be % of clusters passing filter and 3.64 and 3.54 %reads aligned to PhiX spike in.

PS. Analysis can be done using the TruSeq Amplicon BaseSpace App or MiSeq Reporter TruSeq Amplicon Workflow (on the MiSeq).

[larangela]

Thank you for your responses! Those figures are from a secondary analysis derived from SAV. Our primary analysis is done by a qualified third party.

[nucacidhunter]

I wonder if you have run data through MiSeq Reporter as well and checked design manifest for accuracy of target loci.