AISAR (AI SAmpling with NMR Recall selection)

Scripts and datasets (2KOB, CDK2AP1 and GLuc) used in Huang, Ramelot, Spaman, Kobayashi, & Montelione (2025) "Hidden Structural States of Proteins Revealed by Conformer Selection"

AISAR is a computational–experimental framework for identifying fast-exchanging conformational states from NMR data. Unlike conventional NMR methods that rely on spatial restraints, AISAR combines AI-driven conformational sampling of realistic models with Bayesian-like scoring against NOESY and other NMR observables. Applied to Gaussia luciferase, AISAR reveals two interconverting states involving large rearrangements of two lids, binding pockets, and cryptic surface cavities. AISAR also identifies two distinct conformational states of the human tumor suppressor Cyclin-Dependent Kinase 2-Associated Protein 1, demonstrating its utility across diverse protein scaffolds. Validation using the NOESY Double Recall method shows that these multistate models account for more NOESY peaks than individual conformers, supporting the presence of fast-exchanging structural states in dynamic equilibrium. AISAR enables detection and evaluation of conformational heterogeneity and cryptic pockets not resolved by conventional single-state NMR analysis, providing new insights into protein structural dynamics and function.

📘 Documentation

The full AISAR documentation is available in the docs/ folder:

🔧 Software Requirements

R packages used:

bio3d
Documentation and installation:
http://thegrantlab.org/bio3d_v2/tutorials/installing-bio3d
MUSCLE is required for sequence alignment. On macOS (Homebrew):
brew install brewsci/bio/muscle
cluster
DescTools
RColorBrewer

The workflow can be run in RStudio, which is the recommended environment for convenient script editing, visualization, and interactive analysis.

Python Dependencies

numpy
pandas

🌐 External Tools / Webservers Needed

📂 Repository Structure

AISAR/
├─ docs/        # Documentation  (Quick Start, Scripts, Data)
├─ scripts/     # Python scripts used for conformer selection
├─ 2KOB/        # Data and models for 2KOB
├─ CDK2AP1/     # Data and models for CDK2AP1
└─ GLuc/        # Data and models for GLuc

📥 Download pre-calculated AFsample models

2KOB: https://zenodo.org/records/17917670
Unzip into: 2KOB/data_runs/ESmodels/
CDK2AP1: https://zenodo.org/records/15015917
Unzip into: CDK2AP1/data_runs/ESmodels/
GLuc: https://zenodo.org/records/17917708
Unzip into: GLuc/data_runs/ESmodels/

License

RPIBioinformatics/AISAR

About

Resources

License

Stars

Watchers

Forks

Releases

Packages 0

Contributors 2

Languages

Packages