Didong Li

StarTrail identifies rapidly changing regions and cliff genes in spatial omics data, improving boundary detection and uncovering missed biological insights.

The paper introduces hypothesis tests using LLM-derived gene embeddings to improve statistical power in genomics studies.

This paper introduces BGOB, a deep learning framework for longitudinal microbiome interpolation, improving analysis of microbial community trajectories and revealing disease-related associations.

This paper introduces a Riemannian diffusion kernel method for continuous structural connectivity on cortical surfaces, showing improved density estimation compared to spherical heat kernels.

This work optimizes experimental design for spatial genomics studies by proposing structured batch experimental design approaches for atlas construction and region localization.

The paper introduces a new metric, absolute winding number, to differentiate spatial navigation strategies in mice with different APOE alleles, revealing genotype and sex differences linked to Alzheimer's disease risk.

This paper introduces a probabilistic model to align spatial genomics and histology data across different scales and technologies, enabling accurate downstream analysis.

The paper develops hierarchical Gaussian processes and mixture of experts models to analyze EHR data from COVID-19 patients, showing improved predictive performance for clinical markers like albumin.