The software is available at CELLEX: https://github.com/perslab/CELLEX CELLECT: https://github.com/perslab/CELLECT Reproducible code accompanying the paper: https://github.com/perslab/timshel-bmicelltypes … 2/12
-
-
Prikaži ovu nit
-
The growing number of large-scale scRNA-seq atlases (
@humancellatlas) provide a unique opportunity to systematically uncover cell types underlying traits and disease – when properly integrated with GWAS. We built CELLECT to do exactly this. 3/12Prikaži ovu nit -
Most SNPs identified by GWAS are non-coding, suggesting that gene regulation plays an important role. CELLECT is built on the hypothesis that disease-associated SNPs can be linked with the genes they regulate to identify likely etiologic cell-types expressing these genes. 4/12
Prikaži ovu nit -
A key challenge when integrating GWAS and scRNA-seq data is how to robustly represent a cell-type’s unique gene expression profile. To address this challenge, we developed CELLEX. 5/12
Prikaži ovu nit -
CELLEX computes cell type expression specificity profiles. It employs a "wisdom of the crowd"-approach by integrating multiple expression specificity (ES) metrics to capture multiple aspects of ES. We show that CELLEX ES is more robust than individual ES metrics. 6/12pic.twitter.com/c8oMH2DQnk
Prikaži ovu nit -
CELLEX example on Tabula Muris data: the cell types with the highest CELLEX expression specificity for apolipoprotein and glucagon, are hepatocytes and pancreatic alpha-cells, respectively. 7/12pic.twitter.com/LbhPpYrsIB
Prikaži ovu nit -
CELLECT leverages existing ‘genetic prioritization’ methods (e.g. S-LDSC) but makes them much faster and easier to setup and run, thanks to a scalable and reproducible snakemake workflow. 8/12
Prikaži ovu nit -
We applied CELLECT and CELLEX analysis to obesity. Here’s the output of CELLCT (http://mousebrain.org dataset). You can easily do the same analysis for any other GWAS trait. 9/12pic.twitter.com/bvVX6a0f6L
Prikaži ovu nit -
Our results highlight the brain as the master regulator of body weight. Together our results provide an expanded view of the brain’s role in obesity and suggest new avenues for obesity research. Check out the accompanying tweetorial by
@tunepers! 10/12pic.twitter.com/oqE72G6xJw
Prikaži ovu nit -
Bonus: in the Supplementary Notes, we attempted to unify the CELLECT cell type prioritization model with the so-called omnigenic hypothesis proposed by
@jkpritch. 11/12pic.twitter.com/hKlcdognUZ
Prikaži ovu nit -
There are many people I want to thank:
@tunepers for supervising my PhD.@diegoisworking for pioneering RolyPoly. H. Finucane and S. Gazal for S-LDSC support. Christiaan de Leeuw for MAGMA support. Jonathan Thompson,@TobiAlegbe,@StanniusT, Ben Nielsen and@dylanrausch. 12/12Prikaži ovu nit
Kraj razgovora
Novi razgovor -
Čini se da učitavanje traje već neko vrijeme.
Twitter je možda preopterećen ili ima kratkotrajnih poteškoća u radu. Pokušajte ponovno ili potražite dodatne informacije u odjeljku Status Twittera.