PICNIC2-LOCAL and PICNIC2-GLOBAL

Introduction

This README provides instructions for using PICNIC2-LOCAL and PICNIC2-GLOBAL, two 3D ResNets designed to refine AlphaFold tertiary structures. These tools are aimed at enhancing the accuracy of protein structure predictions by leveraging machine learning techniques.

PICNIC2-LOCAL

Description

PICNIC2-LOCAL refines protein tertiary structure one atom at a time. It only runs on the CASP15 input models in the local/casp15_af_models/ directory, as it requires MASS2 and LAW features, which are already generated for those models.

Usage

To refine a protein structure using PICNIC2-LOCAL, follow these steps:

1. Ensure that you use a given input model from a CASP15 target in the local/casp15_af_models/ directory.
2. Run the refinement script using the following command:

   python run_PICNIC2-LOCAL.py local/casp15_af_models/{target_name}/{model_name}
   

   Replace local/casp15_af_models/{target_name}/{model_name} with the path to the specific AlphaFold model you want to refine. An example model is located at local/casp15_af_models/T1104/af2-standard_T1104_1
3. The refined structure will be generated and saved in the local/out_casp15/ directory.

PICNIC2-GLOBAL

Description

PICNIC2-GLOBAL refines the entire protein tertiary structure at once. Unlike PICNIC2-LOCAL, it does not have specific requirements for input model.

Usage

To refine a protein structure using PICNIC2-GLOBAL, follow these steps:

1. Ensure that you have the protein tertiary structure file (in PDB format) ready. You can use the provided example file named sample.pdb.
2. Run the refinement script using the following command:

   python run_PICNIC2-GLOBAL.py {path_to_pdb}
   

   Replace {path_to_pdb} with the path to your input protein structure file. For an example, use sample.pdb
3. The refined structure will be generated and saved in the same directory as the input file.

Note

Both PICNIC2-LOCAL and PICNIC2-GLOBAL utilize advanced machine learning techniques to refine protein structures. It's essential to have the necessary dependencies and libraries installed before running the refinement scripts. See conda_env_packages.txt for the packages needed in a conda environment. Additionally, please ensure that you have sufficient computational resources available to perform the refinement process effectively.