Guanghua Xiao

SpatialOmicsLab provides an integrated environment with AI tools for spatial transcriptomics analysis, outperforming expert workflows in multiple tasks.

This paper introduces a supervised contrastive learning framework for multi-omics prediction models, showing improved accuracy over single-omics approaches in type 2 diabetes and high-fat diet studies.

GPSNorm fits slide, spatial technical and biological effects in one Bayesian negative binomial model, carrying normalization uncertainty into differential expression and lowering log-fold change error in simulations.

This study introduces ReMiTT, a computational method to identify T cell migration trails in spatial transcriptomics data, revealing chemokine trends and enriched pathways associated with T cell movement in tumors.

BiGER is a Bayesian rank aggregation method for meta-analysis of gene lists, handling mixed rankings and missing data with high accuracy.

This paper introduces MambaWSI, a state space model for whole slide image classification that extracts multi-scale local and global features, achieving better performance than existing methods.

This paper uses large language models to extract prognostic histologic features from Ewing sarcoma pathology reports, identifying NSE and S100 as significant biomarkers.

SpaFun identifies domain-representative genes in spatial transcriptomics data using functional principal component analysis, outperforming existing methods in accuracy and efficiency.

This paper introduces Spa3D, a method that reconstructs 3D spatial structures from 2D spatial transcriptomics data, improving spatial domain identification, cell-cell communication analysis, and developmental trajectory modeling.

MisTIC corrects transcript misassignment in spatial transcriptomics data, improving cell type identification, differential expression analysis, and RNA localization.

CytoOne is a probabilistic framework for CyTOF data that integrates batch correction, differential analysis, and visualization, effectively modeling sparse and noisy data.

This study shows that liver cancer initiation depends on metabolic zonation but is not linked to premalignant clonal expansion.

This paper introduces MIDI, a model that predicts cancer drug response and explains drug-target interactions using an attention mechanism, outperforming existing methods in prediction and interpretability.

N-Power AI automates sample size and power analysis in clinical trials using an LLM-based framework, achieving perfect accuracy compared to statistical software.

BayeSMART improves multi-sample spatial clustering by integrating AI-reconstructed single-cell data with spatial context, outperforming existing methods in accuracy and efficiency.

The paper introduces a Bayesian framework to map cell-cell interactions from spatial transcriptomics data, revealing new biological insights in cancer and immune responses.

The paper establishes a database of lung cancer mouse models to improve data comparison and translational relevance by aligning mouse mutations with human lung cancer mutations.

MetaNorm improves normalization of NanoString nCounter data by incorporating meta-analytic priors, enhancing stability and computational efficiency.

This paper introduces BayesDeep, a Bayesian model that reconstructs spatial transcriptomics data at single-cell resolution using histology images and SRT datasets.

ScopeViewer provides a browser-based tool for visualizing spatial transcriptomics data with features for interactive exploration, dual viewing, optimized bandwidth, and data security.

The paper introduces iIMPACT, a method that integrates histology images and molecular data to improve clustering and domain detection in spatial transcriptomics.

STAr provides an integrated platform with curated datasets and methods to evaluate and develop spatial transcriptomics analysis tools.

This study compares neuroendocrine (NE) features in SCLC and NBL, finding similar molecular profiles and shared therapeutic responses but no significant survival association with NE scores.

The paper presents a pan-cancer PDX histology image repository with genomic and pathological annotations for deep learning analysis of tumor regions and xenograft-transplant lymphoproliferative disorder prediction.

The paper introduces a deep-learning method called TPS to quantify zonal gene expression patterns in tissues like the liver, showing it can detect differences between strains, ages, and disease states.

MB-SupCon integrates microbiome and metabolomics data to improve prediction accuracy in microbiome-only datasets, achieving high accuracy in predicting insulin resistance, sex, and race.

This paper develops Sprod, a method to denoise spatial transcriptomics data using location and image information, improving downstream analyses.

MetaPrism enables joint analysis of taxa and gene abundances in metagenomic data to identify potential biomarkers for therapeutic response prediction.

MB-GAN generates synthetic microbiome data that mimic real datasets, preserving key statistical properties for method development.