OpenFold3-preview

Welcome to the Documentation for OpenFold3-preview, a biological structure prediction model based on DeepMind’s AlphaFold3. Developed under a fully open source (Apache 2) license.

Quick-Start Guide

1. Install OpenFold3 using our pip package, more details here

pip install openfold3 

2. Setup your installation of OpenFold3 and download parameters with our script:

setup_openfold

3. Make your first prediction. Note that you may need to configure environment variables.

run_openfold predict --query-json=examples/example_inference_inputs/query_ubiquitin.json

Features

OpenFold3-preview replicates the input features described in the AlphaFold3 publication, as well as batch job support and efficient kernel-accelerated inference.

A summary of the features supported include:

• Structure prediction of standard and non-canonical protein, RNA, and DNA chains, and small molecules

• Pipelines for generating MSAs using the ColabFold server or using JackHMMER / hhblits following the AlphaFold3 protocol

• Structure templates for protein monomers

• Kernel acceleration through cuEquivariance and DeepSpeed4Science kernels - more details here

• Support for multi-query jobs with distributed predictions across multiple GPUs

• Custom settings for memory constrained GPU resources

and more features to come…