ESMARConf2023: {amanida} tutorial

Описание: Presenter: Maria Llambrich
Authors: Llambrich, Maria; Satorra, Pau; Correig, Eudald; Gumà, Josep; Brezmes, Jesús; Tebé, Cristian; Cumeras, Raquel
Session: Tutorials 7
Title: Easy-Amanida tutorial: metabolomics meta-analysis shiny app
Abstract: Metabolomics has emerged as an important field for medical research. To achieve the use of academic findings in routine clinics, robust findings are needed. That is why they are usually evaluated through meta-analysis, the statistical combination, analysis, and evaluation of results from different studies that try to answer or solve the same scientific question. Unfortunately, metabolomics studies rarely disclose all the information needed to perform the gold-standard meta-analysis forest plots, which needs either the raw data or compounds concentrations/intensities, including the standard deviation or standard error disclosed. In fact, the majority of metabolomics clinical studies only report the p-values and fold-changes/trends of the metabolites. Therefore, there is a need to achieve a methodology to do metabolomics meta-analysis with studies that only disclose overall results. So, we have developed Amanida, the first meta-analysis approach that uses the overall statistical results to create a tailored meta-analysis solution for metabolomics. In Amanida, p-values are combined via a weighted Fisher’s method, and fold-changes are combined by average and weighted by the study size (N) after the methodology skewness is reduced. Our solution is available as an R package at CRAN, and we have also created a dedicated website (https://ubidi.shinyapps.io/easy-amani.... In this tutorial we present the shiny app Easy-Amanida, a web for directly perform meta-analysis with overall results. It includes several visualization options: a volcano plot for quantitative results, a vote plot for total behaviors trend (up/down regulation) for each compound, and a explore plot which evaluates the robustness of the vote-counting results with the number of times a compound is found up or downregulated.
GitHub repository: https://github.com/mariallr/amanida

Links: https://esmarconf.org/ESMARConf2023Ad...