Discrepancy between Qbit quant & read count - possible contamination?

[afkoeppel]

We are performing MinION sequencing of DNA extracted from fresh water samples.
We got some preliminary results that suggested possible contamination, so in our most recent round of sequencing we also sequenced a reagent blank from the DNA extraction kit to see if we could rule out contamination from that step.

We first used a Qbit to quant the DNA from each sample, and were encouraged by the result (DNA in the reagent blank was below the detectable threshold).

We then prepped and sequenced our samples (and the blank) using the ONT Rapid Barcoding Kit 96, and found about 5500 reads containing the barcode assigned to the blank. This was a small fraction of the read counts we were seeing for our samples. We had other un-used barcodes that had reads, but never more than a dozen or so.

Is it possible that we somehow introduced contamination during the prep?
We're struggling with how to explain why the Qbit showed no DNA, but the sequencing resulted 5500 reads for the same sample (they mostly classified as E. coli and human when run through the WIMP pipeline).

Has anyone encountered this before? Any advice would be appreciated.

[Anmar06]

Have you added DCS to your samples during library prep? The human component could be contamination but the E.coli reads might be alignments to the DCS (lambda DNA).

[batterprizure]

Hello Anmar06, I have the same question!
backrooms game

[afkoeppel]

Thanks for the suggestion.

For the library prep itself, we only added the reagents in the ONT Rapid sequencing Kit (following the ONT protocol). We did not include the lambda DNA when running our samples.

We had run the lambda DNA prep and sequencing when we first set up the MinION, but on an entirely different flow cell from any of our samples.