Accelerating drug discovery with AI: Insights from Isomorphic Labs

In this episode of Gradient Dissent, Isomorphic Labs Chief AI Officer Max Jaderberg, and Chief Technology Officer Sergei Yakneen join our host Lukas Biewald to discuss the advancements in biotech and drug discovery being unlocked with machine learning.
🎙 Listen on Apple Podcasts: wandb.me/apple-podcasts
🎙 Listen on Spotify: wandb.me/spotify
With backgrounds in advanced AI research at DeepMind, Max and Sergei offer their unique insights into the challenges and successes of applying AI in a complex field like biotechnology. They share their journey at Isomorphic Labs, a company dedicated to revolutionizing drug discovery with AI. In this episode, they discuss the transformative impact of deep learning on the drug development process and Isomorphic Labs' strategy to innovate from molecular design to clinical trials.
You’ll come away with valuable insights into the challenges of applying AI in biotech, the role of AI in streamlining the drug discovery pipeline, and peer into the future of AI-driven solutions in healthcare.
✅ Subscribe to Weights & Biases → bit.ly/45BCkYz
⏳Timestamps:
00:00 Episode introduction and guest overview
05:42 Max's transition from DeepMind to Isomorphic Labs
12:37 Sergei's tech background and move to healthcare
18:54 Early challenges at Isomorphic Labs
25:58 Integrating AI into drug discovery
32:16 Impact of machine learning on drug design
39:07 Introducing AI to drug discovery teams
47:29 AI's role in predicting drug effects
54:55 Future prospects of AI in biotech
01:01:00 AI ethics in healthcare
01:05:00 Broader AI applications in healthcare
01:09:00 Reflections on AI's impact on medicine & wrap-up
🎙 Get our podcasts on these platforms:
Apple Podcasts: wandb.me/apple-podcasts
Spotify: wandb.me/spotify
Google: wandb.me/gd_google
TH-cam: wandb.me/youtube
Connect with Sergei Yakneen & Max Jaderberg:
/ maxjaderberg
/ yakneensergei
/ sergeiiakhnin
/ maxjaderberg
Follow Weights & Biases:
/ weights_biases
/ wandb

[@NachuanShan]

Also 22:08 commenting on Lukas' question. The data in biological world are different from NLP or CV data in various ways, just to name a few:
1. In biology, the experiment data is only an estimation of the physical ground truth and often inconsistent, whereas in many other domain basically the test corpus used for model training is the same in training and real world. So the intrinsic noise within would impact the ceiling of how a model could be evaluated. Since the data is not ground truth, there is a greater gap between model output and reality, given even if the model is perfect on the testing data.
2. The lack of data is real. Partially because bio data is expensive. For CV an annotator could label a dozen or even a hundred pictures per hour and it costs less than $100. But in bio world, on average a single row of data could cost $100-$1000, even over $10k or more for things like protein structure, and takes days or weeks generate. It also requires high level expertise to conduct these experiments, and often repeats need to be done to analyze the intrinsic variances of these data.
3. The format of bio data is so diverse. For LLM, text is all you need, add voice and moving pictures we can train SORA. But in biology, there are hundreds of tasks, structure, affinity, stability, toxicity... each task has many different experiment types.
Well. If you are interested in more about this my twitter is also NachuanShan. I work at BioMap as a data product manager, building protein language models.

[@ayoubalkarim9058]

As mentioned in a previous comment, a significant challenge in applying machine learning to drug discovery projects lies in the scarcity of robust and well-structured data. For instance, a major factor contributing to the failure of drug discovery endeavours is the suboptimal ADMET (absorption, distribution, metabolism, excretion, and toxicity) properties. The landscape could be transformed if we could develop models capable of predicting the outcomes of in vitro assays, allowing us to streamline the selection of well-optimized candidates for pre-clinical trials. However, the publicly available ADMET data is notably deficient in both quality and quantity, leading to the development of models that lack robustness.

[@JagdeepPaniAtEarth]

Excellent questions by Lucas. Insighful discussion.

[@Yogesh-rg1if]

.. becoming comfortable with being uncomfortable ❤️

[@MindFieldMusic]

Great conversation. Love this topic.

[@eliosb989]

Great questions 👍

[@dhruvpatel4948]

Very insightful and informative

[@FLORIDIANMILLIONAIRE]

Where is the data going to come from ? There are strict hippa regulations especially in USA

[@goldnutter412]

The brain is not the mind
The brain is not the mind
The brain is not the mind
Demis gonna win #DeepMind #EZ