Jie Xia

This study uses AI-driven drug design to discover a new MsbA-targeted antibiotic, Y-11, effective against carbapenem-resistant Acinetobacter baumannii with reduced toxicity and resistance.

This paper introduces GEO-BERT, a deep learning model for molecular property prediction, validated through benchmarking and prospective studies on DYRK1A inhibitors.

This paper introduces H-MoE, a hierarchical MoE model for molecular representation learning that improves generalization by incorporating scaffold-level routing and curriculum learning.

This paper develops a machine learning model for plasma protein binding prediction with high accuracy and interpretability, validated on diverse compounds.

The paper discovers a new class of DNA gyrase inhibitors, 1,4-dicarbonylthiosemicarbazides, with potent activity against MRSA through design, synthesis, and biological evaluation.

This paper identifies a novel DYRK1A inhibitor, compound Q17, through computational screening and validates its efficacy in reducing AD-related pathology in vitro and in vivo.

The paper introduces a new benchmark, MUBDsyn, for virtual screening that uses deep reinforcement learning to control biases, showing improved performance over existing benchmarks.

AutoMolDesigner is an open-source AI tool for automated small-molecule antibiotic design, combining generative deep learning and automated machine learning for efficient antibiotic discovery.