oil trading robot for @Shell, predicted the 2008 financial crisis, husband and dad.
Austin, Texas
Joined January 2013
- Tweets72.2K
- Following4.6K
- Followers2.9K
- Likes129K
quant tu peux pas te payer Claude design du coup tu demandes a ChatGPT de one-shot les visualisations de ton paper
LOTUS: lineage-aware optimal transport reconstructs continuous clonal fate dynamics biorxiv.org/content/10.64898…
La machine à vapeur et les vaccins aussi
the main challenge of my PhD for me is to learn to stick to one idea and actually follow through with it instead of jumping left and right every two seconds
JEFE 🇪🇸🇪🇸🇪🇸 retweeted
Every parent wants an honest child. So we built HYPE-PGT, the first genetic predictor of embryo-screening founders.
Kids with a lower HYPE-PGT report more honestly, bin less and understand statistics more. With 10 embryos, picking the lowest-risk one cuts overclaiming by 74%.
JEFE 🇪🇸🇪🇸🇪🇸 retweeted
Excited to share Pycea, a new Python toolkit for lineage analysis and visualization developed by @WilliamNColgan!
Pycea brings lineage trees into the scverse ecosystem, enabling seamless analysis of lineage, gene expression, and spatial data.
biorxiv.org/content/10.64898…
JEFE 🇪🇸🇪🇸🇪🇸 retweeted
It’s more of a stretch to say the CIA killed Lumumba than it is to say they killed JFK tbh
JEFE 🇪🇸🇪🇸🇪🇸 retweeted
In new work, we report a surprising finding: prompting Evo 2 with a handful of sequences of shared function can steer the model to generate new sequences with similar function, a form of generative in-context learning (ICL). This is an emergent phenomenon that we did not anticipate during training.
In a first study, led by Alex Hao, we show that ICL enables few-shot design of mammalian regulatory elements, including novel enhancers with cell-type specificity for a neuronal subtype in the mouse brain.
In a second study, led by @garykbrixi and Alex Hao, we extend these results to the computational design of natural proteins, de novo antibodies, protein-RNA systems, and complete bacteriophage genomes.
As part of this work, we also release several useful new artifacts, including a highly efficient new inference stack. 1/n
