Catch me if you can…
Posts by Nils Homer
Your freezer is lying to you. 🥶 We found frozen samples picked up way more artifactual C>T mutations than DNA stored at room temp. The data surprised us. 🧬👇
It's amazing to me that libdeflate has been around for ~a decade and there hasn't been a production-ready JNI wrapper for it in all that time. So this week I built jlibdeflate: github.com/fulcrumgenom...
Totally agree something has shifted. We used Claude Code extensively for fgumi and it genuinely changed our velocity on the port from Scala to Rust. The key was having deep domain expertise to guide it and catch the subtle stuff. AI as a force multiplier for experienced devs is very real now.
I'm not unique in getting here. 3 days into my effort, @robp.bsky.social wrote about using AI to rewrite piscem (combine-lab.github.io/blog/2026/02...). Then @fulcrumgenomics.com released fgumi this week (blog.fulcrumgenomics.com/p/introducin...).
Something has clearly shifted in AI capabilities..
Super excited to be launching two things today: #RustQC 🦀🧬 and rewrites.bio 🚀
I used AI to rewrite 15 RNA-seq QC tools into a single Rust binary (I've never written any Rust). It ended up being over 60x faster. Here's the story 🧵
seqeralabs.github.io/RustQC/
They’re complementary though: ref-solver uses MD5s from BAM headers when present (very refGet-flavored thinking). We also have a Rust refGet+SeqCol implementation: github.com/fulcrumgenomics/refget-rs
refGet/SeqCol answer “give me the sequence for this digest.” ref-solver answers the reverse: “I have a BAM header, which of the 15+ GRCh38/hg19 flavors is this?” No refGet server can do that lookup from partial dict info.
If you use fgbio or UMIs at all, you should start using fgumi. Up to 100x faster, and soon 2x faster sort than samtools, it’s been a labor of love and something we’ve wanted to do for a long time.
github.com/fulcrumgenom... very much WIP
xsra, in my review, outputs sequences (FASTQ/FASTA/BINSEQ), while the tool I have developed, outputs alignments (SAM/BAM), thus avoiding having to re-align the data when the alignments are available (my use case). So important non-overlapping purposes :)
Went looking for a full-stack GA4GH refget implementation in Rust. Couldn't find one, so I had Claude build it. 🦀 refget-rs server, client, data models, storage, & CLI for Sequences v2.0.0 + Sequence Collections v1.0.0. Serve, query, compare, compute digests. 🔗 github.com/fulcrumgenomics/refget-rs
Last week's fun "how quickly can I solve this" project: github.com/fulcrumgenom... - a small toolkit to i) generate more easily machine parseable kraken2 reports, and ii) filter fastqs based on kraken2 classifications.
"AGBT meet your new bioinformatics team" with hand-drawn portraits of Nils Homer and Tim Fennell
@tfenne.bsky.social and @nilshomer.com are on the ground at #AGBT2026 this week.
Big data. Tight timelines. Real decisions.
#Bioinformatics is where complexity becomes something teams can actually use.
If you’re here, let’s connect.
So long Scala, thanks for all the fish.
You have my email... I could be... convinced...
10th Anniversary Core Values blog series: Valuing People
Valuing People is one of our deepest commitments: to ourselves, to our team, and to our clients. Life is messy—just like biology. When life gets complicated, we don’t look away. We step in.
Read More: blog.fulcrumgenomics.com/p/core-value...
@plasmidsaurus.bsky.social next time in Boston for #ASHG25, consider catering with Bagelsaurus!
Ever want a live code review of your #bioinformatics code?
DM me if you’re at #ASHG25 and I’ll take look to give you suggestions on how to improve reproducibility, maintainability, and overall engineering quality.
10th Anniversary Core Values blog series: Openness
Openness is an action. It’s a positive, and proactive, choice to lean-in.
It begins with a simple but powerful question: What do I know that others need to know?
Read more: blog.fulcrumgenomics.com/p/core-value...
I'll be there to prove I'm not an AI. Lets meet up!
@nilshomer.com, Yossi Farjoun, and @tfenne.bsky.social will be attending #ASHG25 in Boston this year on behalf of Fulcrum Genomics. Let us know if you'll be there too!
@nilshomer.com, Yossi Farjoun and I will be attending ASHG in Boston this year on behalf of Fulcrum Genomics. Let us know if you'll be there too!
The Charles river doesn’t have trees… just saying.
Here’s to another ten years working with awesome and talented people on challenging and impactful Bioinformatics problems!
blog.fulcrumgenomics.com/p/ten-years-...
We love working with people who notice how things fit together, who make unwieldy problems work better. The 'Meet the Team' series spotlights humans behind our bioinformatics work. We hope you'll feel like you've met us — or want to. Meet Erin McAuley: blog.fulcrumgenomics.com/p/meet-erin-...
Remember, to be generous also means accepting help to allow others the opportunity to be generous. You cannot give that which you do not have.
Rather than feeling unworthy or like a burden, I choose to be grateful for the time they shared, giving them the chance to be generous and reaffirming my own value.
I regularly volunteer my time meet with folks in the field. Instead, today reached out for guidance and help, and I was met with generosity as someone took the time to meet with me.
Saying no when someone offers to help deprives them of the opportunity to be generous.