Back to feed
AI in healthcarePrimary source

NVIDIA and healthcare partners unveil protein-design model and develop AI for cancer care

NVIDIA and the UW Medicine Institute for Protein Design introduced the open RFD4-Proteina model for designing and searching proteins. The company also announced work with Siemens Healthineers on adaptive radiotherapy and with life sciences firms developing research agents on its BioNeMo platform.

Preferred on Google
Text size

Image: nvidianews.nvidia.com · source articleEditorial excerpt for news reporting

At GTC Berlin, NVIDIA announced collaborations with healthcare and research organizations working on protein design, drug discovery and cancer care. One of the announcements was RFD4-Proteina, an open model developed with the UW Medicine Institute for Protein Design.

The model generates protein designs for specified goals and lets researchers search and refine candidates using molecular inputs, constraints and scoring objectives. In testing described in a research draft, it outperformed the evaluated baselines in computational design success and generation throughput across four benchmark suites: protein binding, small-molecule binding, enzyme scaffolding and DNA binding.

RFD4-Proteina is available through Amazon Bio Discovery, Anthropic’s Claude Science, Benchling, Databricks, Google Cloud, Hugging Face, Nebius, OpenAI and Tamarind Bio. Laboratory experiments are needed to establish how selected designs perform under tested conditions.

Siemens Healthineers develops adaptive radiotherapy

Siemens Healthineers and NVIDIA are working on adaptive radiotherapy that combines agent-assisted planning, patient-specific simulation and real-time computing. The companies are exploring NVIDIA Nemotron open models to help oncologists prepare and refine treatment plans.

The proposed workflow also includes dose calculations accelerated by NVIDIA CUDA and the NVIDIA Holoscan platform, which can connect live sensor data with AI and simulation. Siemens Healthineers said the collaboration builds on earlier work, including its Accela system; the companies did not announce a timetable for new solutions.

Scientific agents for drug discovery

Automata, Benchling, Biolevate, Boltz, Danaher, Hologen and Therna Biosciences are using NVIDIA’s BioNeMo platform to develop agents that connect drug design, laboratory processes and biological research. Automata, for example, is linking agents to automated lab infrastructure, while Boltz is developing tools to predict molecular structures and assess candidates.

Genedata is developing an AI co-scientist to coordinate protein design and testing in Danaher’s laboratory. The workflow will incorporate RFD4-Proteina, with experimental results informing the next design cycle. Hologen is working on tools to analyze medical imaging and clinical data, while Therna Biosciences is connecting experimental evidence with RNA medicine design.

Other healthcare uses for AI

NVIDIA also said Avelios, Corti, Dedalus, Dawn Health and Sword are building products on Nemotron, ranging from AI scribes to personalized patient support. PolyAI is using the NVIDIA Agent Toolkit in a multilingual dialog service with medical software provider Corilus, whose network spans more than 70,000 healthcare professionals in Belgium and France.

What we know

  • NVIDIA and the UW Medicine Institute for Protein Design introduced the open RFD4-Proteina model.
  • The model outperformed evaluated baselines for computational design success and generation throughput across four benchmark suites.
  • RFD4-Proteina is available through nine partner services, including Amazon Bio Discovery, Google Cloud and OpenAI.
  • Siemens Healthineers and NVIDIA are exploring AI-assisted planning, patient-specific simulation and real-time computing for adaptive radiotherapy.
  • Corilus has a network spanning more than 70,000 healthcare professionals in Belgium and France.

Trust: Primary source

  • The story cites a primary statement. It establishes what its author said, but does not guarantee every claim is true.
  • Sources: NVIDIA Newsroom.
Related news
View sources1

COMMUNITY

Discussion

—

Loading comments…