Machine Learning in Materials Science

Reconstructing living materials as a computable design space with multi-agent reasoning

This paper introduces LiveMat, a framework that transforms literature into a computable design space for living materials, improving reasoning across domains.

An SE(3)-equivariant Crystal Structure Prediction Framework for Prospective Identification and Development of Bioactive Nucleoside Self-Assembling Materials

The paper introduces SE3CSP, an SE(3)-equivariant deep-learning framework for crystal structure prediction of nucleosides, achieving 57% accuracy and enabling the discovery of bioactive self-assembling materials.

CGAgentX: Agentic AI Framework to Develop Transferable Coarse-Grained Models

CGAgentX develops transferable coarse-grained models using an agentic AI framework with specialized agents that autonomously optimize parameters to match experimental and simulation data.

Enhancing non-local interaction modeling for ab initio biomolecular calculations and simulations with ViSNet-PIMA

This paper introduces ViSNet-PIMA, a model that improves non-local interaction modeling in biomolecular simulations, outperforming existing MLFFs and reducing calculation errors in AI2BMD.

"Visualize, describe, compare" - nanoinformatics approaches for material-omics

The paper introduces a computational package for visualizing, describing, and comparing nanoparticle datasets from super-resolution microscopy to analyze material properties and heterogeneity.

Benchmarking Universal Machine Learning Force Fields with Hydrogen-Bonding Cooperativity

This study benchmarks three MLFF models on their ability to capture hydrogen-bond cooperativity in N-methylacetamide polymers, finding MACE-OFF most accurate.

Automated Discovery of Therapeutic Biomaterial for Renally Impaired Hyperuricemia Patients by Natural Language Processing and Machine Learning

The paper develops MatWISE, a framework using NLP and ML to discover a biomaterial for renally impaired hyperuricemia, identifying δ-MnO2 as a promising UA-lowering agent.