Genomica

Genomica is an R package that provides a suite for the analysis of functional orthologs from the KEGG Orthology (KO) database. Genomica is based on a linear mixed model approach, whilst also allowing to test for interactions between maximum two predictors, therefore outputting both a list of significant, FDR adjusted KOs across the treatment layout and the results of the enrichment analysis based on the LMM-informed enriched or depleted orthologs, both in the different treatment groups and cumulatively.

Only two data frames are needed as input to run Genomica, one containing the data (KO abundance) and one containing metadata information (e.g., treatment layout), moreover if required, Genomica will carry out the log10 transformation of pre-normalised data. The outputs from Genomica are summarised both in (tab-delimited) .txt and in .xlsx files describing the whole list of statistical results per feature and a list of significant comparisons, respectively. Moreover, an Enrichment directory is created within the output directory, which contains the results (.txt, .xlsx and .tiff) from the enrichment analysis.

The tests carried out in Genomica develop through linear mixed models (LMM) via calling the package lmer4 (Bates et al., 2015) and calculating the P-value via type III ANOVA using the Satterthwaite's method through the package “lmerTest” (Kuznetsova et al., 2017). Moreover, the tests, and ultimately the significance levels are based on the calculation of the false discovery rate (type I error) under repeated testing. This is carried out via calling the package fuzzySim (A. Marcia Barbosal, 2015), which uses the Benjamini & Hochberg correction to generate p.adjust values based on the p values generated by the LMM. Finally, the enrichment analysis is performed via calling MicrobiomeProfiler(Yu G., Chen M, 2024), which, through the analyses performed in Genomica also requires enrichplot (Yu G., 2025) and clusterProfiler (Xu, S. et al., 2024, Yu G, 2012).

If used to analyse other types of complex datasets instead of KOs (e.g., AMR tables or quantitative gene tables from qPCR experiments), Genomica will only return the lists of significant comparisons across the predictors, as established by the user, without performing the enrichment analysis.

To download and use Genomica, please follow the instructions via this link: sgalg/Genomica