Accompanied by a less polished nextflow pipeline to handle multiple samples and pre-processing - github.com/allydunham/d.... Configurably downsample, merge and trim reads before quantification plus FastQC, SeqKit Stats and a quanitification QC plot. Inspired by the github.com/cancerit/QUA....

I have more features planned to support my work, for instance multiple constructs in one library, combinations of sub-libraries and a region type expecting variants to a base sequence, but issues and pull-requests with other bugs and suggestions are very welcome too.

Simulated tests and benchmarks and our datasets suggest the tool is generally accurate and robust as well as pretty quick. Feels like a state that could be useful more widely so a good time to share, although I do expect more bugs to come out with wider usage!

Currently position in reads, flanking patterns and full alignment can be used to extract regions and extact matching, hamming distance and (bounded) Levenshtein used to compare to your library. This lets you nicely balance match accuracy with speed for your design.

It takes in (paired) fastq/a files, a JSON expected read structure, an optional expected combinations TSV and a strategy for extracting and matching variable regions and outputs count tables against your library.