Yeah, this was a real headache, and was the impetus for writing the pairwise program. Alignment success rates also varied based on chemistry (see v9 vs. v10 in supplement). Maybe because some of our reads were shorter than medaka was intended for?

We are satisfied with the results and will use this pipeline until such a time that 100% accurate single molecule Nanopore is achieved. We are now using this in many of our synthetic biology pipelines -- directed evolution, MPRAs, DMS libraries. Hit me up with questions or for help.

Grab serum from virally infected (here, COVID) patients to map and monitor single molecule quasi-species distributions of the circulating genomes (would suggest deeper than MinIon reads, we repeat sample preparation multiple times. GridIon probably best).

Accurately quantify edit distributions from CRISPR/Cas9 genome KO pools.

So what can you do with something like this?

QC your viral preps to detect random-break packaging issues not observable in Illumina sequencing.

We develop:
1. Bootstrap algorithm to benchmark # of clusters needed to give accurate sequences on a per run basis.
2. Our own aligner to overcome tandem repeat issues
3. A Docker-based GUI wrapper for the program.

Get it here:
github.com/JGEnglishLab...

Issue #3, Origin of Mismatches. We see rare mutations in our sequences. Are they real? Yes, we develop a method to PCR the samples directly and confirm by Sanger. Where are they from, bacterial passaging or PCR? Ran LTE of bacteria and repeated PCR -- both contribute, PCR orders of magnitude more.

Issue #2, Abundance Reproducibility. Is the content of a sample accurately reproduced in the data output? We mixed plasmids of known sequence together in a logarithmic range of concentrations and repeatedly measured the sample demonstrating reproducible, accurate readout of abundances.

Issue #1, Chimeric Molecules. Inevitable oligo contamination between tagging and amplification rounds creates hybrid pairs. How to accurately discard them computationally? Using a deep barcoded template as an internal barcode we observed chimera formation patterns and programmed to filter them out.

This work was led by Adam Zahm and Caleb Cranney in our group. There are lots of unique molecular index (UMI) strategies out there, but both the software packages and ground-truth sequence accuracy validations were severely lacking. So we decided to tackle this problem and deliver a unified solution

so you're saying American financial policy decisions should be made to appease and court our least responsible voters? Or that we should pump and dump them like our current President and make them poorer?

"GOP warms to idea children should live" just to keep it wholesome.