Cancer Drug Discovery AI Agentic Workflow R&D

Описание: PDF: https://www.chemicalqdevice.com/cance...
Many Generative AI developments exist for Drug Discovery applications. However, additional methods to further automate the process for more informed drug compound leads is advantageous. Industry experts have identified AI agentic workflows as a key area to solve more complex problems.

As planning and multi-agent collaboration improves, more productive interactions will appear between an LLM and another LLM or smaller machine learning models. In specific, LangChain, LangGraph, or CrewAI libraries provide necessary frameworks to accomplish an increasing number of orchestrated tasks.

Experts such as Andrew Ng, Harrison Chase, and Arsenii Shatokh have conducted recent seminars to update the progression of AI agentic workflows. Research groups such as LIAC’s ‘ChemCrow’ in 2023 and LAMM-MIT’s ‘MechGPT’ in 2024 have utilized chemical AI agentic workflows that can be used analogous to Cancer Drug Discovery applications.

“Instead of you prompting AI and it responds in a few seconds, we now see instances where I can tell it ‘Dear AI, please do research for me and write a report’ and then it goes off for half an hour, and browses the internet, summarizes all these things, comes back in half an hour with a report.” This is not working as good as I just described it, but is working much better now than three months ago.
Andrew Ng on AI's Potential Effect on the Labor Force | WSJ regarding Automation. February 2024.    • Andrew Ng on AI's Potential Effect on the ...