Exponentiating Single-Cell Sequencing

on YouTube ↗35m

" summary

Generates a comprehensive single-cell lung atlas and pairs it with transcriptome-wide spatial profiling using Nanostring WTA of post-mortem COVID-19 lungs to map diffuse alveolar damage stages and microenvironments that distinguish early (exudative) from late (organised) DAD and to remap spatial pathology biomarkers into the single-cell reference to nominate therapeutic targets. Implements sample multiplexing with Muxlet to demultiplex pooled donors, sequences 1.2 million cells (and 1.6M in follow-up), achieves ~5–20× throughput (100k droplet-resolved cells), reports ~$600 per sample, and introduces statistical tool Memento for mean/variance/covariance testing with ~30% power gain over pseudobulk.

AI-written summary · about

" chapters (24)

1. 0:00Intro
2. 1:10Exponential increase in single-cell genomics throughput
3. 2:14Massively-parallel single-cell RNA-seq
4. 4:36Need for population scale single cell sequencing
5. 6:36Sample multiplexed scRNA-seq
6. 10:12SLE Immune Cell Census
7. 13:16CD4 decrease explains lymphopenia
8. 14:07Monocytes mediate bulk expression changes
9. 15:21Decomposing gene expression for eQTL mapping
10. 17:39Cell-type-specific cis-eQTLs enriched for CRE Cell-type-specific
11. 18:15CTS EQTLs enriched for immune GWAS hits
12. 19:19Public sharing of data
13. 19:30Other applications of mux-seq (mostly blood)
14. 21:33Poisson statistics limits droplet-based sc-seq
15. 22:39SCITO-seq: combinatorial indexed cytometry sequencing
16. 23:48XYzeq: spatially resolved single-cell sequencing
17. 27:30Solution: Memento
18. 28:07Memento overview
19. 28:49Part 1: Estimation - generalization/simulation
20. 29:48Part 1: Estimation - FISH Validation
21. 30:37Part 2: Efficient resampling using unique counts
22. 31:10Part 2: Memento calibration and reproducibility
23. 31:37App: eQTL discovery - better ATAC enrichment
24. 31:58Summary