Peptone at NVIDIA GTC 2026: Tackling the "Undruggable" Proteome

I'm proud to share that two key members of our team will be presenting at this year's NVIDIA GPU Technology Conference:

🎤 Carlo Fisicaro, CTO
🎤 Michele Invernizzi, Sr. Scientist

Their session — "Accelerating Small Molecule Drug Discovery for Hard-to-Target Proteins" — dives into how we're pushing the boundaries of computational drug discovery for intrinsically disordered proteins (IDPs), the "undruggable" targets that conventional structure prediction tools struggle with.

What you'll learn

• How we built a flow-matching sequence-to-ensemble model for disordered proteins, trained with augmented data and experimental biophysical constraints

• Our dedicated benchmark suite for evaluating ensemble predictors — and diagnosing where single-structure tools break down

• How GPU-accelerated ensemble prediction pipelines plug directly into protein structure generation workflows for hard-to-target oncology, CNS, and immunology programs

This work sits at the intersection of generative AI, molecular simulation, and drug discovery — powered by NVIDIA's DGX Platform, BioNeMo, and Hopper architecture.

If you're at GTC, don't miss session S82028. Come say hi to Carlo and Michele!

👉 Register here: https://register.nvidia.com/flow/nvidia/gtc26/ap/page/catalog/session/1765632901542001teKm