Multivariate analysis of Pleurodeles waltl injection outcomes using FAMD FactoMineR (RStudio 4.3.2) used to probe sources of variability during an Adeno-associated viral (AAV) screen

In this study, we analyzed the efficiency of different adeno-associated viral (AAV) injections in transfecting neurons in the salamander Pleurodeles waltl. To query sources of variation in AAV injection outcomes, we analyzed metadata and outcomes for an additional 68 intraparenchymal injections performed 39 in post-metamorphic Pleurodeles salamanders. This dataset included both quantitative variables (age, weight, viral genomes (v.g.) injected, and a cell transduction score ranging from 0 to 4, see STAR Methods) and categorical variables (here referred to as qualitative variables: serotype, promoter, reporter, single vs. dual injection, manufacturer, and injection site). To assess the associations between these variables, we performed a Factor Analysis for Mixed Data (FAMD), a principal component method that is designed to determine significant sources of variability within datasets that contain both quantitative and qualitative data types. This dataset contains the injection outcomes and metadata for AAV injections administered in the salamander Pleurodeles waltl. The RactoMineR package (https://CRAN.R-project.org/package=FactoMineR) was used to determine significant contributions of a number of variables contributing to injection outcomes. Analysis of these data revealed that, among other factors, age co-varies with injection score. Therefore, we conclude that increased animal age decreases the efficacy of this tool.