Developing new computational tools for spatial transcriptomics data

Project Summary
Spatial transcriptomics is a groundbreaking new technology that allows measurement of gene ac-
tivity in a tissue sample while mapping where the activity is occurring. It holds the promise to facilitate
our understanding of spatial heterogeneity underlying essential phenotypes and diseases, such as
neurodegenerative diseases and cancer. However, the development of bioinformatics infrastructures
and computational tools has fallen seriously behind the technological advances. The lack of proper
computational approaches presents current data analysis barriers that significantly hinder biological
investigations. The overarching goal of this proposal is to address some of the most pressing ana-
lytic challenges facing profiling and interpreting spatial transcriptomics data, including 1) lack of robust
identification of genes with spatial expression patterns across a variety of technical platforms, 2) lack
of tools to identify structures, microenvironments as well as developmental trajectory on the tissue,
and 3) lack of tools that can jointly analyze spatial transcriptomic data across multiple samples and
multiple data sources. In the proposal, we will work on the following aims: Aim 1. Develop nonpara-
metric tools for identifying genes with spatial expression patterns. Aim 2. Develop spatially aware
dimension reduction tools for detecting structures and developmental trajectories on the tissue. Aim 3.
Develop integrative association tools for spatial transcriptomic analysis across multiple samples and
datasets. All the methods will be implemented in user-friendly software and disseminated to the sci-
entific community. Successful achievement of all aims will dramatically increase the power of spatial
transcriptomics analysis, and facilitate the application of these cutting-edge technologies to transla-
tional and clinical studies.

Grant Number: 5R01HG011883-04
NIH Institute/Center: NIH
Principal Investigator: Mengjie Chen