Bioinformatics Seminar
Time: 11AM
Venue: Davis Auditorium and Online
9 June 2026
Computational methods and workflows for sample-level analyses of spatial transcriptomics data
Lukas WeberBoston University
Spatial transcriptomics enables the measurement of up to transcriptome-scale gene expression together with spatial coordinates within tissue sections. Unsupervised statistical methods are widely used within computational analysis workflows for these data, for example for analyses such as feature selection and clustering. While initial methods were typically designed to analyze data from individual tissue samples, recent methodological developments have focused on integrated analyses of data from multiple samples. Motivated by collaborative studies of spatial transcriptomics data from the dorsolateral prefrontal cortex and locus coeruleus brain regions in postmortem human brain samples, we have developed several new methods for analyses of these data. In this talk, we will discuss methods for feature selection to identify spatially variable genes (nnSVG), sample-level clustering summary metrics (spatial domain boundary density metric), and quality control. These methods are available as R packages within the Bioconductor framework. In addition, we will discuss an extensive, community-driven resource consisting of an online book containing interactive R code and example datasets demonstrating spatial transcriptomics analysis workflows (Orchestrating Spatial Transcriptomics Analysis with Bioconductor).