Submission Tilte
Machine Learning in Structural Bioinformatics: Predicting Protein Dynamics and Functional States
Submission Abstract:
This thematic issue aims to explore the cutting-edge intersection of structural bioinformatics and machine learning, with a specific focus on predicting protein dynamics and alternative functional states. While tools like AlphaFold2 have made significant strides in static protein structure prediction, they still struggle to capture dynamic aspects like conformational transitions, intermediate states, and allosteric effects, which are vital for understanding protein function and druggability.
The proposed issue will bring together state-of-the-art methodologies including stochastic MSA subsampling, Markov State Models (MSMs), metadynamics simulations, and hybrid approaches integrating molecular dynamics (MD) with deep learning. Articles will cover both method development and practical applications, such as the identification of cryptic pockets, analysis of CDR loop dynamics in antibodies and nanobodies, and conformational landscapes in disease-relevant targets.
We will encourage submissions that use integrative multi-omics, functional annotations, and ensemble-based simulations to improve protein function prediction and precision drug design. The issue also welcomes comprehensive reviews, benchmarking studies, and case studies combining experimental validation (e.g., NMR, cryo-EM, or X-ray) with computational predictions.
This thematic issue will serve as a timely and impactful resource for computational biologists, AI researchers, and structural biologists seeking to understand the full spectrum of protein behavior through innovative, data-driven approaches.