Xiaoning Qian

CARTnet: Template-guided B-cell Epitope Prediction with Contact-aware Attention

CARTnet improves B-cell epitope prediction using a contact-aware deep learning framework, outperforming existing methods on internal and external datasets.

Agent K: An Agent-Orchestrated Pipeline for Mapping Genes to Biosynthetic Pathway towards Metabolite Discovery

Agent K automates the mapping of bacterial genomes to biosynthetic pathways and metabolite discovery using an agent-orchestrated framework with a large-language model.

UPLiFT: Uncertainty Parameterized Activation for Lightweight Fine-Tuning

UPLiFT improves parameter efficiency in fine-tuning protein language models by optimizing activation space instead of network weights, achieving competitive performance with fewer trainable parameters.

Structure-Based TCR-pMHC Binding Prediction and Generalization to Unseen Peptides

This work evaluates factors affecting generalization of GNN-based TCR-pMHC binding predictors to unseen peptides and proposes architectural improvements.

A Synthetic Microbial Therapy Rewires Antitumor Immunity Across Multiple Cancer Types

This paper introduces a synthetic microbial therapy that enhances antitumor immunity across multiple cancer types in mice, showing durable responses and synergy with chemotherapy.

An Integrated Computational Antigen Discovery Pipeline with Hierarchical Filtering for Emerging Viral Variants

This paper introduces a computational pipeline for antigen discovery that uses filtering and machine learning to identify conserved neutralizing epitopes in viruses like SARS-CoV-2 and RVFV.

Rapid sequence-based screening of structure-disrupting protein mutations

The paper develops a sequence-based method to identify structure-disrupting mutations using protein language models, reducing the need for expensive structure prediction.

Interpretable transcriptome-to-phenotype modeling of cell-painting nuclear morphology features from RNA-seq under low-dose radiation exposure

The paper models how low-dose radiation affects nuclear morphology by linking RNA-seq data to cell-painting imaging features over time using an interpretable inverse modeling framework.

An Integrated and Configurable End-to-End Pipeline for Longitudinal Cell Painting Analysis

The paper introduces SCALE, a pipeline for longitudinal cell painting analysis, demonstrating its ability to handle imaging variability and enable consistent comparisons across time points.

PhenoGraph: A Multi-Agent Framework for Phenotype-driven Discovery in Spatial Transcriptomics Data Augmented with Knowledge Graphs

PhenoGraph automates phenotype-driven analysis of spatial transcriptomics data using a multi-agent system and biological knowledge graphs to improve interpretability and efficiency.

MUSTANG: MUlti-sample Spatial Transcriptomics data ANalysis with cross-sample transcriptional similarity Guidance

MUSTANG performs multi-sample spatial transcriptomics analysis by integrating cross-sample similarity and spatial correlation to improve cellular deconvolution.

Biologically Interpretable VAE with Supervision for Transcriptomics Data Under Ordinal Perturbations

The paper introduces EXPORT, a VAE framework for analyzing ordinally perturbed transcriptomics data, enhancing interpretability by incorporating pathway knowledge and ordinal constraints.

Sensitivity analysis of genome-scale metabolic flux prediction

The paper develops a sensitivity analysis strategy for genome-scale metabolic flux prediction using TRIMER, showing its utility in guiding optimal experimental design for metabolic engineering.

TRIMER: Transcription Regulation Integrated with MEtabolic Regulation

TRIMER integrates transcriptional and metabolic regulation to predict metabolic fluxes and gene states, outperforming existing methods on simulated and experimental data.