From f3f8288b127ad39bba3ecee3ab0a102318363d66 Mon Sep 17 00:00:00 2001 From: Marc Subirana Granes <48090783+msubirana@users.noreply.github.com> Date: Wed, 27 Nov 2024 08:04:21 -0700 Subject: [PATCH] Update content/06.a-gene-module-perspective-for-genetic-studies.md Co-authored-by: Milton Pividori --- content/06.a-gene-module-perspective-for-genetic-studies.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/content/06.a-gene-module-perspective-for-genetic-studies.md b/content/06.a-gene-module-perspective-for-genetic-studies.md index 5b28d7e..0d2c57a 100644 --- a/content/06.a-gene-module-perspective-for-genetic-studies.md +++ b/content/06.a-gene-module-perspective-for-genetic-studies.md @@ -5,7 +5,7 @@ Each panel shows a component of the PhenoPLIER framework [@doi:10.1038/s41467-023-41057-4]. *(a)* First, latent variables (LVs) or gene modules are learned from transcriptome data using the Pathway-Level Information Extractor (PLIER) matrix factorization method. PLIER generates matrix $\mathbf{Z}$, which has gene weights for each module, and matrix $\mathbf{B}$, which has the samples in the latent space. -*(b)* Schematic illustrating how gene-trait associations from Transcriptome-Wide Association Studies (TWAS) and gene-drug scores from LINCS L1000 are projected into the latent space for joint analysis. This involves a matrix multiplication that transforms a traits × genes matrix into a traits × LVs matrix, enabling integration and interpretation at the gene module level. +*(b)* Schematic illustrating how gene-trait associations from TWAS and gene-drug scores from LINCS L1000 are projected into the latent space for joint analysis. This involves a matrix multiplication that transforms a traits × genes matrix into a traits × LVs matrix, enabling integration and interpretation at the gene module level. *(c)* Schematic of a gene module-based drug reporposing framework, where the projection of TWAS and LINCS L1000 data is used to compute a drug-disease score. *(d)* Schematic of a regression model that tests whether genes that belong to a module (using a column of $\mathbf{Z}$) tend to be more strongly associated with a trait (using $p$-values from TWAS). *(e)* (top) Example of a gene module identified as LV246 analyzed in [@doi:10.1038/s41467-023-41057-4].