NVIDIA Clara Discovery Platform
NVIDIA 2020 GTC Artificial Intelligence Assisted Drug Development
Eco
Nvidia’s entry into the medical field. From the cover we can see a ring structure:
Natural Language -> Viral Gene Sequence -> Viral Protein Structure -> New Drug Docking with Viral Protein -> Molecular Docking Dynamics Simulation -> Clinical Trial of Drug Candidate Detecting Medical Images of Patient’s Lungs -> Natural Language
Introduction
At the Nvidia GTC 2020 Fall Conference, Nvidia announced the NVIDIA CLARA DISCOVERY drug development framework, which aims to accelerate the progress of global drug development. With the news of Nvidia’s acquisition of ARM before, I realized that Nvidia has the ambition to rule the edge computing field in the future. Not surprisingly, this GTC conference Nvidia updated many product lines, including large to data center computing module BlueField-2, small to edge field jetson Nano 2G, the most exciting for me is still ** CLARA DISCOVERY**, because my own short-term goal is how to use edge computing in the medical field, such as individuals can calculate the current status of their own diseases through edge computing devices combined with their own diagnostic data obtained from hospital examinations, as well as judgmental detection of Covid-19 CT, ChestXRay.
What is it
Clara Discovery integrates many frameworks and applications, including gene sequence analysis, protein docking and structure prediction, molecular dynamics simulation, medical image recognition and segmentation, and natural language processing.
Gene sequence generation
NVIDIA Clara Parabricks Collection
Image processing
Protein structure
Molecular dynamics simulations
Highlights
NVIDIA Clara Imaging is divided into two main parts:
Clara Train SDK -> for training the model
Architecture diagram:
Getting started with Clara Train SDK
Clara Deploy Bootstrap -> Deploy the model to the kubernets cluster
Architecture diagram:
