Preeti
The article discusses the potential of Large Language Models (LLMs) in early drug discovery, particularly focusing on how AI-led biotechs can leverage them. LLMs, based on deep learning models, have gained popularity in various industries due to their ability to process vast amounts of data. In the context of drug discovery, companies like Recursion and Insilico Medicine have made announcements related to LLMs. The article explains how LLMs work, highlighting their training on large text datasets to understand human language.
Challenges such as LLM hallucinations are addressed, emphasizing the importance of curating data to avoid unrealistic outcomes. The article also categorizes users of LLMs into two types: those who use existing models and those who build their own. The former approach is discussed in detail, with a focus on using interactive systems like ChatGPT.
Limitations of LLMs are acknowledged, with a comparison to other models like Diffusion Models (DMs) used in AlphaFold 3. The article mentions upcoming articles in the series that will delve into practical examples of using LLMs in science. The author, Dr. Raminderpal Singh, is a key opinion leader in the techbio industry with extensive experience in computational modeling systems. He leads the HitchhikersAI.org community and is the CEO of Incubate Bio, providing services to life sciences companies. His background includes leadership roles at IBM Research and Eagle Genomics Ltd, with a PhD in semiconductor modeling and numerous patents and publications.
Source link
Source link: https://www.drugtargetreview.com/article/151039/part-one-what-can-scientists-do-with-llms-today/
GIPHY App Key not set. Please check settings