Computational approaches enhance the design of molecular glue degraders for undruggable proteins.

Dec 8, 2025Drug discovery today

Using AI to Design New Drugs for Hard-to-Treat Proteins

AI simplified

Abstract

Many proteins are difficult to target with traditional drugs because they don’t have clear binding sites. Molecular glue degraders (MGDs) are a new type of drug that can help target these tough proteins. However, only a few of these drugs have been approved for use, and there are still challenges like side effects and resistance. This paper discusses how recent advancements in computer-based design and artificial intelligence are making it easier and more accurate to create these drugs. Future developments in technology could lead to new treatments for diseases that are currently hard to treat.

AI simplified

Most proteins remain 'undruggable' by traditional approaches, which are unable to engage targets because of a lack of well-defined binding pockets, causing a bottleneck in drug discovery. Molecular glue degraders (MGDs) have emerged as a promising therapeutic strategy for targeting previously undruggable proteins. However, despite their potential, only a few MGDs have received FDA approval, highlighting gaps in off-target effects, drug resistance, and substrate availability. Here, we discuss recent MGD breakthroughs driven by the integration of structure-based computational approaches and AI platforms, which have accelerated MGD design with improved accuracy. Looking ahead, advances in quantum computing and AI-based generative models might open pathways to innovative treatments, targeting diseases once considered incurable.

Full Text

Full text is available at the source.

View full text (XML) ↗View full text ↗

Abstract

Full Text

what lands in your inbox each week: