Find open-source science resources

Improves simultaneous inference under dependence of tests by estimating a collapsed null distribution through resampling. Accounting for the dependence between tests increases the power while reducing the variability of the false discovery proportion. This dependence is common in genomics applications, e.g. when combining flow cytometry measurements with microbiome sequence counts.

Combine ideas of log-linear analysis of contingency table, flexible response function estimation and empirical Bayes dispersion estimation for explorative visualization of microbiome datasets. The package includes unconstrained as well as constrained analysis. In addition, diagnostic plot to detect lack of fit are available.

MMUPHin is an R package for meta-analysis tasks of microbiome cohorts. It has function interfaces for: a) covariate-controlled batch- and cohort effect adjustment, b) meta-analysis differential abundance testing, c) meta-analysis unsupervised discrete structure (clustering) discovery, and d) meta-analysis unsupervised continuous structure discovery.

Utilities for microbiome analysis.

MaAsLin 3 refines and extends generalized multivariate linear models for meta-omicron association discovery. It finds abundance and prevalence associations between microbiome meta-omics features and complex metadata in population-scale epidemiological studies. The software includes multiple analysis methods (including support for multiple covariates, repeated measures, and ordered predictors), filtering, normalization, and transform options to customize analysis for your specific study.

MaAsLin2 is comprehensive R package for efficiently determining multivariable association between clinical metadata and microbial meta'omic features. MaAsLin2 relies on general linear models to accommodate most modern epidemiological study designs, including cross-sectional and longitudinal, and offers a variety of data exploration, normalization, and transformation methods. MaAsLin2 is the next generation of MaAsLin.

This explorative ordination method combines quasi-likelihood estimation, compositional regression models and latent variable models for integrative visualization of several omics datasets. Both unconstrained and constrained integration are available. The results are shown as interpretable, compositional multiplots.

Starting from a microbiome dataset (16S or WMS with absolute count values) it is possible to perform several analysis to assess the performances of many differential abundance detection methods. A basic and standardized version of the main differential abundance analysis methods is supplied but the user can also add his method to the benchmark. The analyses focus on 4 main aspects: i) the goodness of fit of each method's distributional assumptions on the observed count data, ii) the ability to control the false discovery rate, iii) the within and between method concordances, iv) the truthfulness of the findings if any apriori knowledge is given. Several graphical functions are available for result visualization.