Recent Open Source Work by Fulcrum Genomics

Highlights from the last 3-4 months of our work in open source software

At Fulcrum Genomics we rely daily on a wide range of open source software. We also try to be good community members, and contribute to the wealth of open source software bioinformaticians use. This work takes on two distinct flavors:

1. Publishing and maintaining open source libraries and toolkits that we own (e.g. fgbio, fqtk, fqgrep, and more).

2. Actively engaging and solving issues with other libraries, tools, platforms we use instead of working around them.

This post is going to focus on the latter - the work we do outside of our core projects. It’s important to me that we give back to the community, and generally leave things in a better state than we find them. In fact, “Public Benefit” is one of Fulcrum’s core values. While it might be easier and faster, in our day to day work, to just work around problems that we find, we’d rather take the extra time to log issues, open PRs and generally solve problems the right way rather than taking the easy way out. Doing so helps the community, and our future-selves will thank us!

---

Here’s a brief summary of some of the open-source work we contributed to in the first few months of 2025:

• Released a new GitHub action, gha-timer (repo), that makes it easy to get customized and finer-grained timing within GitHub Actions. Along the way we became a GitHub “official partner organization” so we could more easily share the action within the GitHub Marketplace.

• Worked really hard to reduce the CI build times for all PRs in bioconda, culminating in the publishing of another GitHub Action: fetch-through-merge-base. By changing builds from cloning the entire bioconda-recipes git repo (with tens of thousands of commits, and hundreds of active branches) to just checking out the subtree required for the specific branch build, we shaved several minutes off nearly all recipe builds!

• Fixed a problem in picard’s CollectSamErrorMetrics to correctly handle lower-case bases from the reference FASTA.

• Published pybwa, a new set of python bindings for bwa that provide access to both bwa aln and bwa mem functionality directly from python. pybwa is fully type-hinted and returns pysam AlignedSegments that should be familiar to anyone working with BAM files in python.

• Released the setup-latch GitHub Action that sets up the Latch SDK for use in GitHub Actions, allowing deployment of pipelines to Latch from GitHub CI.

• Collaborated with the Samuk Lab at UC Riverside to support the new release of Pixy, a tool that estimates average nucleotide diversity within and between populations.

Check back in a few months for another post with, I hope, an even longer list of community contributions from our team!