NVIDIA Blueprint: Generative Virtual Screening for Drug Discovery

The NVIDIA Blueprints for generative virtual screening demonstrate how generative AI and accelerated NVIDIA NIM microservices can be leveraged to design optimized small molecules more intelligently and efficiently.

In this lab, you will get the ability to see NVIDIA BioNeMo NIM microservices working together to generate meaningful predictions within a small virtual screening workflow will be demonstrated. The aim is to highlight to users how easy it is to query and integrate these tools, as well as the flexibility of such a workflow.

You will be exposed to the following BioNeMo NIM microservices:

• NVIDIA NIM for AlphaFold2
• NVIDIA NIM for DiffDock
• NVIDIA NIM for MolMIM

• To understand how NVIDIA BioNeMo NIM microservices can be brought together to accelerate drug discovery and screening.
• Understand how to access NVIDIA NIM API documentation.

What you will interact with:

• Jupyter Notebook
• NVIDIA NIM for AlphaFold2
• NVIDIA NIM for DiffDock
• NVIDIA NIM for MolMIM

Under the covers, you will be running on:

• Dell Reference Architecture for Generative AI with NVIDIA
• Red Hat OpenShift
• Run:ai