Shaamil Karim
Founder of Atlas Discovery, Predicting human response to drugs in clinical trials
Company
Shaamil Karim is listed as a founder of Atlas Discovery. Predicting human response to drugs in clinical trials
Founder traction evidence
Public posts and activity attributed directly to this founder.
- LinkedIn
Founder says he started Atlas Discovery. Post says Atlas Discovery raised funding from Y Combinator.
Founder says he started Atlas Discovery. Post says Atlas Discovery raised funding from Y Combinator.
- X
We trained a foundation model of patient biology that predicts drug response from a single biopsy before treatment. As a case study, we tested on a phase 3 trial of ustekinumab in IBD and get a 0.76 AUROC, enough to run it with ~450 fewer patients at the same statistical power.
We trained a foundation model of patient biology that predicts drug response from a single biopsy before treatment. As a case study, we tested on a phase 3 trial of ustekinumab in IBD and get a 0.76 AUROC, enough to run it with ~450 fewer patients at the same statistical power.
- X
Excited to share our paper at @icmlconf genbio right now. We ablate the core design choices in latent perturbation prediction and then te...
Excited to share our paper at @icmlconf genbio right now. We ablate the core design choices in latent perturbation prediction and then tested the model on a real-world out-of-distribution target discovery task from Pfizer, achieving SOTA results. Read more from @splicewiring!
- X
Highly recommend this podcast. @BenjamineYLiu does a good job explaining the thesis that drug design is becoming commoditized and drug de...
Highly recommend this podcast. @BenjamineYLiu does a good job explaining the thesis that drug design is becoming commoditized and drug development will be the bottleneck. Casual targets of disease are also an issue, but real-world data can help build better predictive models.
- X
Thank you, @AsteraInstitute , for selecting my essay on virtual cell benchmarks and the need to ground them in clinical outcomes! Congrats to the other awesome essays @pkoo562 @Prashant_Garg_ @_JaeeonLee_ @nivedithasi @MattLeighton5 @ChristinaErnst_ @HarshuMusunuri
Thank you, @AsteraInstitute , for selecting my essay on virtual cell benchmarks and the need to ground them in clinical outcomes! Congrats to the other awesome essays @pkoo562 @Prashant_Garg_ @_JaeeonLee_ @nivedithasi @MattLeighton5 @ChristinaErnst_ @HarshuMusunuri
- X
Everyone in pharma cites the looming patent cliff of $300B between 2025 and 2030 and how it accounts for 1/6 of pharma revenue. So I wa...
Everyone in pharma cites the looming patent cliff of $300B between 2025 and 2030 and how it accounts for 1/6 of pharma revenue. So I wanted to put this in context with other periods in history. The 2010-15 revenue risk was almost double 2025-30 (as a % of total pharma revenue).
- X
I got the chance to write a deep dive on Virtual Cells for @Contrary_Res . It's the longest article written on this topic (50 min reading time), and the goal is to motivate why AI virtual cell models offer a much-needed paradigm shift for the drug discovery industry. Link
I got the chance to write a deep dive on Virtual Cells for @Contrary_Res . It's the longest article written on this topic (50 min reading time), and the goal is to motivate why AI virtual cell models offer a much-needed paradigm shift for the drug discovery industry. Link
- X
And this is a lower bound. One drug, one indication, a model trained on a fraction of the data that exists. The trial has always been our coarsest model of the patient. This is evidence starting to move from the trial into the model.
And this is a lower bound. One drug, one indication, a model trained on a fraction of the data that exists. The trial has always been our coarsest model of the patient. This is evidence starting to move from the trial into the model.