Qiao Liu

scMTG reconstructs single-cell temporal dynamics from unpaired time-series data by learning cell representations and transitions, outperforming existing methods in interpolation and transition matrix inference.

MultiFlow predicts single-cell multiomic perturbation responses by jointly modeling gene expression and chromatin accessibility across cellular contexts.

RegFM improves gene regulation modeling by integrating cis-regulatory elements and trans-acting regulators in human cells, outperforming existing methods in multiple tasks.

scDiffusion-X generates high-fidelity multi-omics data and enables modality translation with uncertainty quantification and gene regulatory network inference.

CREATE identifies cell-type-specific cis-regulatory elements using a deep learning model that integrates genomic sequences, chromatin accessibility, and chromatin interactions.

This paper introduces a method to infer gene regulatory networks using scATAC-seq data from gene perturbation experiments, showing improved accuracy and interpretability in GRN reconstruction.

This study introduces epiBrainLLM, a framework using genomic LLMs to improve causal analysis of genotype-brain-clinical pathways in Alzheimer's disease.

Deconer evaluates reference-based cell type deconvolution methods using simulated and real gene expression data, comparing 14 methods across multiple criteria and providing guidance for users.

EpiGePT is a transformer-based model that predicts epigenomic signals and chromatin contacts by incorporating transcription factor activities and 3D genome interactions, outperforming existing methods in cross-cell-type prediction.

DeepDrug uses graph-based deep learning to predict drug-drug and drug-target interactions, outperforming existing methods and identifying potential SARS-CoV-2 treatments.

This paper introduces scDEC-Hi-C, a deep generative framework for analyzing single cell Hi-C data, showing improved clustering and imputation over existing methods.

A NeurIPS 2021 benchmarking competition evaluated over 2600 submissions from 280 competitors across three multimodal single-cell integration tasks using a purpose-built benchmarking dataset and cloud evaluation pipeline.

DeepCAGE predicts chromatin accessible regions using deep learning, integrating sequence data and transcription factor activity, and identifies disease-associated variants.

OpenAnnotate provides a web server for efficiently annotating chromatin accessibility of genomic regions across biosamples, tissues, and systems.

DeepCDR predicts cancer drug response by integrating multi-omics data and drug chemical structures using a hybrid graph convolutional network, outperforming existing methods.

scDEC uses deep generative networks to simultaneously model and cluster scATAC-seq data, showing better performance than existing tools and enabling trajectory inference and cell type discovery.