AlphaFold 3 is great - but it still needs human help to get chemistry right
Briefly

Recent AI advancements, particularly with tools like AlphaFold 3, are transforming how scientists predict biomolecular structures. However, inherent limitations in handling stereochemistry challenge their practical applications.
While AlphaFold 3 has shown remarkable capabilities in structure prediction, its current limitations in accurately modeling stereochemistry are becoming increasingly evident, raising concerns for its reliability in complex biological contexts.
Read at Nature
[
|
]