Aaron Wenteler
@aaron_wtr
PhD AI for Drug Discovery @QMUL, @EPFL, @MSDintheUK. ex-{@TUDelft, @CureVacRNA}. Researching the intersection of AI and biology to improve human health. 🧬🤖
1/14 Excited to introduce PertEval-scFM! It provides benchmark and evaluation tools for perturbation effect prediction models, including single-cell foundation models (scFMs). Paper and GitHub link at the end of the thread! 🧵👇
University budgets everywhere are getting slashed, and we hear many PhD students with accepted ICML papers can no longer afford to attend. We are stepping in to offer travel grants for researchers who lost funding. Sponsored by Pillar VC, Ormoni Bio, Latent Labs, and Cimulate
The Machine Learning community moved to Bluesky, and I haven't seen a place with so many ML/AI technical conversations since Twitter 2021. For all I know, these same conversations might be happening here in X, but the algorithm decided not to show them to me anymore.
I’d love to hear biologists weigh in on the fundamental question: can you predict the impact of cell perturbations better by studying the natural variation in a population of healthy cells (more data), or by studying cells that have been perturbed genetically or chemically?
virtual cells are currently bottlenecked by compute, not novel data: drug discovery is an iterative search process (design, test, analyze) through therapeutic design space guided by a dynamics model directly trawling this therapeutic space with large hypothesis-free…
Had an amazing first month in 🇨🇭. Now gearing up to travel to ICML. Shoot me message if you’ll be around in Vancouver. Would love to connect and to talk anything AI, drug discovery and beyond 😎🧬
Congrats @pdhsu and team, very impressive results in here! Looking forward diving into this in more detail
Today @arcinstitute releases State, our first perturbation prediction AI model and an important step towards our goal of a virtual cell State is designed to learn how to shift cells between states (e.g. “diseased” to “healthy”) using drugs, cytokines, or genetic perturbations
How is AI driving drug discovery? Listen to @aaron_wtr as he explains how #AI and #MachineLearning can sift through the data to refine drug targets. Read more in #CPPrecisionmedicine: AI approaches for the discovery and validation of drug targets ➡️bit.ly/3XlAWpL
Deep-learning at scale is so complicated engineering-wise that assessing the value of a new idea without an army of top-notch colleagues to help you is IMO impossible.
Proud to share that PertEval-scFM got accepted into ICML! Hope to see you all in Vancouver. 🇨🇦
1/14 Excited to introduce PertEval-scFM! It provides benchmark and evaluation tools for perturbation effect prediction models, including single-cell foundation models (scFMs). Paper and GitHub link at the end of the thread! 🧵👇
Congrats @NadigAjay on TRADE out now in @NatureGenet: nature.com/articles/s4158… These statistical metrics enable more meaningful comparisons in Perturb-seq atlases. Also, now find the HepG2 and Jurkat Perturb-seq datasets on GEO GSE264667!
🔬 How well do single-cell foundation models (scFMs) predict perturbation effects? This study introduces PertEval-scFM, a benchmark evaluating zero-shot scFM embeddings against simpler baseline models to assess their effectiveness in predicting transcriptional responses to…
It was a pleasure talking about our recent single-cell foundation model benchmark, PertEval-scFM, at the Multiomics Reading Group at @Mila_Quebec. Many thanks to the organizers for the invitation and to @valence_ai for sharing the talk youtu.be/DCezfwQkkAE?si…
It was a pleasure talking about our recent single-cell foundation model benchmark, PertEval-scFM, at the Multiomics Reading Group at @Mila_Quebec. Many thanks to the organizers for the invitation and to @valence_ai for sharing the talk youtu.be/DCezfwQkkAE?si…
I spoke with @Nature about the transformative potential of foundation models for biology—particularly in personalized medicine and clinical diagnostics. Take a look at the technologies to watch in 2025. An exciting year ahead in AI, biology and medicine! nature.com/articles/d4158…
1️⃣ "PertEval-scFM: Benchmarking Single-Cell Foundation Models for Perturbation Effect Prediction” @aaron_wtr @martiinaina biorxiv.org/content/10.110…
A central mistake in biology was to name genes. This over-simplification made reconciling what is happening on the molecular level a mess - it's not rare to find reports of opposite mechanisms in different contexts, claimed involvement in dozens if not hundreds of different…
Very proud to have been part of the GenPlasmid team led by Will. Great write-up on the project 👇
Over the past week, hundreds of folks came together to participate in Evolved 2024, a Bio x ML hackathon hosted by @Lux_Capital, @EvoscaleAI and @envedabio Our team won 3rd-place🥉 and the @Polaris_HQ new dataset challenge! Here's our project, GenPlasmid... 🧵