Dong Si

The paper introduces a large dataset of cryo-EM RNA motifs and benchmarks machine learning models for structure prediction.

ContrastQA improves protein complex structure quality assessment using label-guided graph contrastive learning, achieving lower ranking losses than existing methods on CASP16 datasets.

GOBoost improves protein function prediction by addressing long-tail distribution of Gene Ontology Terms using an ensemble strategy and specialized loss functions.

AnglesRefine refines 3D protein structures using a Transformer model focused on torsion angles, improving model quality and reducing computational time.

The paper develops a system for predicting protein-ligand binding sites from Cryo-EM maps and generates drug-like ligands using a combination of PLBP methods and a de novo structure modeling tool.

This paper introduces DeepTracer-Refine, an automated method to refine AlphaFold2 protein structures, improving accuracy and efficiency compared to existing methods.

The paper introduces DeepTracer-2.0, an AI pipeline that automatically models protein-DNA/RNA complexes from cryo-EM maps, improving accuracy and efficiency in macromolecular structure prediction.

This paper introduces SynthQA, a hierarchical machine learning method for protein model quality assessment that outperforms traditional methods by incorporating multi-scale features and feature relationships.

DeepTracer is a deep learning method for fast cryo-EM structure modeling of protein complexes, showing improved accuracy and efficiency compared to existing methods.

Cascaded CNNs segmenting secondary structure, backbone and Cα voxels, plus tabu-search tracing, place 88.5% of Cα atoms within 3Å across 50 cryo-EM maps versus Phenix's 66.8%.

DeepTracer predicts protein backbone atoms, secondary structures, and amino acid types from cryo-EM density maps with high accuracy and speed.

This paper introduces an automated method for selecting density thresholds in cryo-EM maps to improve backbone structure prediction accuracy.

This paper explores AI methods, including reinforcement learning, for protein structure prediction and discusses their implications for bioinformatics and biotechnology.

The paper introduces AngularQA, an LSTM-based method for protein model quality assessment using a novel representation of atom information, achieving a correlation of 0.651 on CASP12.