Excited to share this preprint that describes my latest work on using GPUs to accelerate processing of RNA-seq data.
The title says it all: "RNA-seq analysis in seconds using GPUs" now on biorxiv www.biorxiv.org/content/10.6... and github github.com/pachterlab/k...
Figure 1 shows they key result
A centromere but not just a centromere: structure and evolution of a selfish chromosomal supergene in monkeyflowers www.biorxiv.org/content/10.6...
I am happy to announce the launch of a new Arabidopsis genomics resource! Check out arabidopsislyrata.org Now you can easily look at the natural genetic variation across the entire species range of A. lyrata and A. arenosa.
Why did I use Codex? Because MSU does not currently allow us to use Claude Code for research endeavors like this (no enterprise version available).
I know a lot of folk on here are going to be immediately resistant to incorporating AI into work this way. I am not really sure what to tell you. That said, I would be happy to hear from anyone who wants to test the agent on their own data and let me know how it goes.
I then worked with the agent to run through two very different RNA-seq data sets from my lab. It did what appeared to be a great job and I was able to update its skill set after each run through with what we had learned along the way. Each update was then easily pushed onto GitHub.
By the weekend, I had to try it for myself. Why not? I downloaded Codex and logged in with my MSU enterprise version. An hour later, I had an agent that could run through the full WGCNA analysis.
Sitting at the Department of Energy Annual PI meeting last week in DC, I saw a lot of talks about using Claude Code and Codex to develop agents for various bioinformatics workflows. I also listened to a number of podcasts on the subject.
For context, I spent weeks trying to implement WGCNA as a postdoc only to eventually give up after crashing it too many times. Last year, I used Claude to help me power through the development of code to implement WGCNA, but it still took a week or so.
Weighted Gene Co-expression Network Analysis (WGCNA) has been the go to program for analysis of gene expression modules for nearly two decades but it is a difficult to implement. To make it easier to use WGCNA, I used Codex to build a human in the loop agentic workflow. github.com/lowrylab/WGC...
Are you a grad student attending #Evol2026? Submit your talk for the SSE Hamilton Award for Outstanding Graduate Student Presentation! Finalists will present their talks during the virtual portion of the meeting. Apply during registration. Deadline April 15! shorturl.at/qDIP4 @evolmtg.bsky.social
The next BAPG has been announced! April 11 at Davis organized by the inimitable @jrossibarra.bsky.social! bapg-conference.github.io
Nice new collaborative paper out this week led by Rob Heckman, using a data set we collected way back when I was a postdoc: "Correlational selection and genetic architecture shape the evolution of the leaf economics spectrum in a perennial grass" www.pnas.org/doi/10.1073/...
The Guerrero Lab at NC State University seeks a postdoc in computational evolutionary genetics. Ideal candidates have quantitative skills. Apply by March 10, 2026, via rfguerre@ncsu.edu. More info: rguerrer.org #postdoc
Hey @derekdenney.bsky.social & Annabelle Taylor wrote a nice essay about why we should care about plant reproduction + heat. ecoevorxiv.org/repository/v...
Crop scientists have long recognized that reproduction is going to be the first point of failure for plants with rising global temperatures. However, this has often been overlooked by those studying natural populations. In this new review, we try to change that. ecoevorxiv.org/repository/v...
The Bart Lab at UC Berkeley is looking for a postdoc. Please share!
Turner et al. investigated the role of chromosomal rearrangements during habitat shifts in two locally adapted populations of Drosophila from the island of São Tomè; findings suggest that rearrangements may act as a source of innovation.
🔗 doi.org/10.1093/gbe/evaf236
#genome #evolution #drosophila
Many Wood Storks on a tree in a swamp.
Dance of the Wood Storks.
🎉🎉🎉
The program for the 2026 New York Population Genetics meeting, hosted by at the @simonsfoundation.org on March 9th 2026, is now up: events.simonsfoundation.org/event/7c91dd....
What happens when the ocean — and the expectations placed on scientists — are changing at the same time? As a Bilinski Fellow at the UC Davis Bodega Marine Laboratory, Mikhaila Redovian explored how marine scientists are responding through collaboration and new ways of working:
At least these grass seem capable of harvesting themselves.
This feels straight out of an intense dream about our switchgrass diversity panel experiments.
Up here, folks are leaf blowing fresh snow.
New preprint from our big collaborative evolution experiment in 9 whole lakes in Alaska, written by McGill grad student @lucaseckert.bsky.social
www.biorxiv.org/content/10.6...
What happens if you put a mixture of multiple source populations together to complete & evolve in multiple new lakes?
This article is now published! academic.oup.com/nargab/artic...
We’ve added a few new analyses. First off, we show that, while gene presence absence variation (PAV) scales with evolutionary distance in both plants and animals, the base level and rate of accrual are both twice as high in plants.
New review!
Theory & a practical guide to structural variants in popgen🧬
Many thanks to my co-authors: @rebekahoomen.bsky.social
@annatigano.bsky.social @marenwellenreuther.bsky.social @janawold.bsky.social @dlfield.bsky.social @clairemerot.bsky.social
onlinelibrary.wiley.com/doi/10.1111/...
PEQG! My favorite conference ever*. Single session, fantastic talks, great discussions. Only once every 2 years. And in the best conference location too! Abstract deadline coming up Feb 5!
*Disclosure: I am required to say it's the best conference because it is and I told my mom I wouldn't lie.
The genetic and developmental enigma of rhizomes: crucial traits with limited understanding. arxiv.org/abs/2601.10847
