Xuequn Shang

DesignMaster: A Multi-Conditional Diffusion Framework for Rational PROTAC Design

DesignMaster improves PROTAC design by incorporating linker length and physicochemical constraints through a diffusion framework, achieving better validity and recovery than existing methods.

Structure-aware deep learning predicts influenza antigenicity and guides vaccine strain recommendation

This paper introduces Vir3D, a structure-aware deep learning model that predicts influenza antigenicity and improves vaccine strain selection by incorporating 3D structural information from amino acid sequences.

A variational sparse Gaussian-process method for detecting spatially variable genes and cellular interactions from spatial transcriptomics

VISGP identifies spatially variable genes and cellular interactions from spatial transcriptomics data, outperforming existing methods in detecting ligand-receptor pairs.

SpatialClaw: A Memory-Augmented Autonomous Ecosystem for Spatial Omics Analysis

SpatialClaw provides a memory-augmented system for spatial omics analysis, enabling end-to-end workflows through natural language interaction and outperforming standard models in benchmarking.

Cell Type-informed Characterization of Spatial Niches from Spatial Multimodal and Multi-omics Data

SpaNECT is a framework for characterizing cell niches from spatial multimodal and multi-omics data, outperforming existing methods in diverse biological contexts.

Pathway-informed Universal Domain Adaptation for Single-cell RNA-seq Data

This paper introduces scPathOT, a pathway-informed domain adaptation method for single-cell RNA-seq data, demonstrating its ability to accurately annotate cell types and uncover functional mechanisms across diverse biological contexts.

FluNexus: a versatile web platform for antigenic prediction and visualization of influenza A viruses

FluNexus provides a web platform for antigenic prediction and visualization of influenza A viruses, offering tools for data preprocessing, prediction, and visualization of antigenic evolution.

LoFT-TCR: A LoRA-based Fine-tuning Framework for TCR-Antigen Binding Prediction

LoFT-TCR improves TCR-antigen binding prediction using LoRA fine-tuning of ESM-2, achieving better accuracy than existing methods.

Efficient Identification of Short Tandem Repeats via Context-Aware Motif Discovery and Ultra-Fast Sequence Alignment

FastSTR efficiently identifies and characterizes tandem repeats across 13 species with high accuracy and speed, uncovering new genomic elements and disease-related variations.

Genome-wide profiling of short tandem repeat somatic instability reveals associations with age, sex and brain-related traits

The paper introduces searchSTR, a tool for genome-wide STR somatic instability analysis, and finds associations between STR instability and age, sex, brain morphology, and neurodegeneration markers.

Efficient integration of spatial omics data for joint domain detection, matching, and alignment with stMSA

This paper introduces stMSA, a deep learning model for integrating spatial omics data, improving domain detection, matching, and alignment across tissue slices.

A Multi-Property Optimizing Generative Adversarial Network for de novo Antimicrobial Peptide Design

This paper introduces MPOGAN, a generative adversarial network for designing antimicrobial peptides with multiple desired properties, showing improved performance in antimicrobial activity and reduced cytotoxicity.

TridentTCR: Predicting T Cell Receptor Specificity for Antigens and Autoimmune-Related Targets via a Topology-Aware Graph and Large Language Model

TridentTCR predicts TCR specificity for antigens and autoimmune targets using a graph neural network and large language model, achieving better classification and identifying cross-reactivity in type 1 diabetes.

A Multimodal Graph Learning Framework for Versatile Spatial Transcriptomics Analysis with SpatialModal

SpatialModal integrates gene expression and histological images in spatial transcriptomics to improve spatial analysis and domain identification.

Spatial geometry-aware deep learning for deciphering tissue structure from spatially resolved transcriptomics

SpatialGEO improves identification of spatial regions in spatial transcriptomics data and reveals biological insights in tumor microenvironments and mouse development.

Deep Learning-Based Classification of CRISPR Loci Using Repeat Sequences

This paper introduces a deep learning method to classify CRISPR-Cas systems using repeat sequences, improving classification accuracy for subtypes with limited data.

DisConST: Deciphering Spatial Domains Using Distribution-aware Contrastive Learning for Spatial Transcriptomics

DisConST improves spatial domain detection in spatial transcriptomics by using distribution-aware contrastive learning, achieving better accuracy than existing methods.

CellAgent: LLM-Driven Multi-Agent Framework for Natural Language-Based Single-Cell Analysis

CellAgent uses an LLM-driven multi-agent framework to automate scRNA-seq and spatial transcriptomics analysis through natural language interactions, achieving 60% efficiency improvement over manual methods.

BERTE：High-precision hierarchical classification of transposable elements by a transfer learning method with BERT pre-trained model and convolutional neural network

BERTE improves TE classification accuracy using BERT and CNN, achieving up to 21% higher F1-score than existing methods.

Comprehensive Hierarchical Classification of Transposable Elements based on Deep Learning

This paper introduces CREATE, a deep learning framework for TE classification, and WebDLTE, a web platform for real-time TE annotation, showing improved performance over existing tools.

A multi-scale feature fusion model based on biological knowledge graph and transformer-encoder for drug-drug interaction prediction

This paper introduces ALG-DDI, a multi-scale feature fusion model for drug-drug interaction prediction that outperforms existing methods by integrating attribute, local biological, and global semantic information.

Flexibility and rigidity index for chromosome packing, flexibility and dynamics analysis

The paper introduces a flexibility-rigidity index (FRI) model to quantify chromosome flexibility using Hi-C data, showing improved accuracy and efficiency compared to Gaussian network model (GNM).

A Deep Learning-based Genome-wide Polygenic Risk Score for Common Diseases Identifies Individuals with Risk

This paper introduces DeepPRS, a deep learning-based polygenic risk score model that outperforms existing methods in predicting risk for common diseases using genome-wide genotype data.

A novel framework for single-cell Hi-C clustering based on graph-convolution-based imputation and two-phase-based feature extraction

This paper introduces ScHiC-Rep, a framework for single-cell Hi-C clustering that improves imputation and feature extraction, outperforming existing methods on human and mouse datasets.

Efficient and scalable integration of single-cell data using domain-adversarial and variational approximation

DAVAE integrates single-cell data across samples, technologies, and modalities by removing batch effects and enabling transfer learning and cell type prediction.

Integrating multi-network topology for gene function prediction using deep neural networks

The paper proposes DeepMNE-CNN, a method that integrates multiple biological networks using semi-supervised autoencoders and convolutional neural networks to improve gene function prediction.

Identifying Emerging Phenomenon in Plant Long Temporal Phenotyping Experiments

This paper introduces TEP-Finder, an algorithm to identify emerging phenomena in plant phenotyping data, showing it finds more functionally specific and biologically significant patterns than existing methods.

Combining Gene Ontology with Deep Neural Networks to Enhance the Clustering of Single Cell RNA-Seq Data

The paper proposes two methods integrating Gene Ontology with neural networks to improve scRNA-seq clustering, showing better performance than existing approaches.

A learning-based framework for miRNA-disease association prediction using neural networks

The paper introduces a neural network framework, MDA-CNN, for predicting miRNA-disease and miRNA-phenotype associations, showing improved performance over existing methods.