Spatial Transcriptomics Identifies Neighbourhoods and Molecular Markers of Alveolar Damage...

on YouTube ↗30m

" summary

Built a 600,000‑cell single‑cell lung atlas from 15 studies and applies NanoString GeoMx (≈18,000 probes) to 233 ROIs from 33 post‑mortem lungs scored as PRES, E‑DAD, M‑DAD and O‑DAD. Deconvolution plus MILO shows loss of alveolar epithelial cells and big expansion of immune cells and fibroblasts; NMF finds recurring tissue niches; random‑forest on ROI profiles reaches ~73% accuracy for damage stage. Key markers include upregulated SRPN1 and CDH11; provides molecular candidates and spatial maps tied to specific pathological stages.

AI-written summary · about

" chapters (25)

1. 0:00Intro
2. 0:16COVID-19 does not need much of an introduction
3. 2:30Fibrosis and hyperinflammation in the lung
4. 3:24Characterization of COVID-19 lung using spatial transcriptomics
5. 4:06Spatial transcriptomics using Nanostring GeoMx platform
6. 5:09Workflow for analyzing Nanostring data
7. 5:54Building a lung reference atlas
8. 6:40Integrated COVID-19 lung reference of healthy and COVID-19
9. 7:07Cell type clusters and marker genes
10. 7:46Differential abundance testing using Milo
11. 9:23Overview of spatial transcriptomics data
12. 10:01Distinct transcriptomic signatures across pathologies
13. 11:19Pathway analysis identifies coagulation as dysregulated
14. 12:15Overview of cell type abundance by pathology
15. 12:54Overview of cell type abundance comparing control and COVID
16. 13:51Summary of cell type abundance changes
17. 14:14Breakdown by pathology state shows progression
18. 14:56Fine grained view of cell abundance changes in COVID-19
19. 16:07Identification of patterns of cellular organization
20. 17:17Niches change as pathology progresses
21. 17:47Evolution of cell states across disease pathology
22. 18:08Use a random forest classifier to identify markers of each pathological state
23. 18:57Identification of genes predictive of different categories
24. 20:19Next-generation pathology for human tissues
25. 21:00Summary and conclusions

" related talks

• 40mGraph neural networks for biomedical insightsRohit Tripathy · PyData NYC 2024
• 32mMulti-Omic Integration for Understanding DiseaseRon Shamir · Computational Challenges in Very Large-Scale 'Omics'
• 24mHigh-throughput cellular analytics to identify central nervous system targets for gene therapyHelen Poldsam · Data Con LA 2020
• 30mLeveraging Molecular Data for Drug DiscoveryHaiyan Huang · Computational Challenges in Very Large-Scale 'Omics'
• 29mMachine Learning Techniques for Omics Data AnalysisVesna Pajic · DSC Europe 23
• 44mNetwork Propagation as a Tool for Deciphering Disease MechanismsRoded Sharan · Network Biology 2016
• 29mHarnessing Multimodal Single-Cell Sequencing Data for Integrative AnalysisElizabeth Purdom · Computational Challenges in Very Large-Scale 'Omics'
• 20mDiagnostic Medical application of MLCarrie Weinberg · MLconf NYC 2019