Enrichment of GWAS signal has provided crucial insights into tissues, cell types and pathways that underlie disease. But what about tissue structures identified from emerging spatial data sets? This is what we wanted to tackle, embedded a world-leading hub for spatial omics at @scilifelab.se and KTH - ThreadSky

tuuliel.bsky.social • 23 days ago

Enrichment of GWAS signal has provided crucial insights into tissues, cell types and pathways that underlie disease. But what about tissue structures identified from emerging spatial data sets? This is what we wanted to tackle, embedded a world-leading hub for spatial omics at @scilifelab.se and KTH

Comments

tuuliel.bsky.social•23 days ago

We processed 31 published 🐭 & 🙂 spatial datasets from 8 organs with genetically implicated (and drug target) genes in 32 human diseases from @opentargets.org. We then mapped the tissue structures where disease genes are higher expressed than the null, using our new approach STEAM.

tuuliel.bsky.social•23 days ago

While genes often showed a broad enrichment across a tissue, in about 20% of disease-tissue pairs only a subset of tissue structures were enriched, highlighting the value of spatial data. Cool patterns across many diseases/tissues - and a big resource for domain specialists to explore!

tuuliel.bsky.social•23 days ago

The brain is particularly cool, with its well-defined substructure. We see the expected neuropsych enrichment in cortex, while Alzheimer’s genes are enriched in white matter, thalamus and choroid plexus. Spatial correlation shows subclusters of disease genes in distinct spatial domains.

tuuliel.bsky.social•23 days ago

But what about spatial vs cell type data? Analyzing sc/sn data from the brain suggests that some diseases could arise from dysfunction in a particular cell type that may or may not be spatially enriched, while others may derive from dysfunction of multiple cell types in a given tissue structure.

tuuliel.bsky.social•23 days ago

Drug target genes for a given disease share enrichments with genetically implicated genes. We also show vignettes of genes that are genetically associated to a disease, expressed in disease-enriched structures, and targets for known drugs.

tuuliel.bsky.social•23 days ago

In summary, we show spatial transcriptomics data provides and important layer of insights for functional interpretation of GWAS data. We hope to see broad use of the tools and resources we compiled, and look forward to even broader applications as the ST data sets improve in scope and resolution.

Comments

Posting Rules

Reply