Table 1.

Overview of some free bioinformatics software for the integration of information across several omics techniques.

Name	Characteristics	Integration of Types of Omics	Type of Analysis	Reference and URL
Cytoscape	standalone software	Mainly protein–protein, protein–DNA and DNA–DNA interactions, but plug-ins (apps) are available for all types of omics	Provides tools to visualize complex molecular and genetic interaction networks, but also network analysis, enrichment analysis, ontology analysis and pathway analysis (e.g., KEGG) is possible.	[33] https://cytoscape.org/
MOFA	R package (via Bioconductor)	All types (multi-omics)	Multi-Omics Factor Analysis enables the unsupervised integration of heterogeneous data sets via a generalization of principal components analysis. MOFA implements hidden factors of biological and technical sources of variability and represents integrated data in an interpretable low-dimensional form.	[34] https://github.com/bioFAM/MOFA
LUCID	R package (via CRAN)	Mainly genomics and metabolomics; integration of phenotypic data	Uses Latent Unknown Clusters with Integrated Data models to distinguish unique genomic, exposure and informative biomarkers or omics effects. Latent underlying relationships with phenotypic traits are estimated in cluster estimations using directed acyclic graphs (DAG). Prediction of phenotypes possible. Visualization of data integration with Sankey diagrams.	[35] https://github.com/USCbiostats/LUCIDus
MultiDataSet	R package (via Bioconductor)	Epigenomics, transcriptomics; assay data, feature data, phenotypic data stored in single object	Does not provide tools for analysis itself, but constructs an R data-storage object that contains multiple data sets, making managing and subsetting multiple and non-complete data sets possible. This data set can be plugged in to other R packages for analysis, for instance for multivariate co-inertia analysis (MCIA in omicade4) or clustering of multiple tables (in iClusterPlus)	[36] https://bioconductor.org/packages/release/bioc/html/MultiDataSet.html
Logicome Profiler	standalone Unix software	Applied to genomics and metagenomics, but applicable to any omics data	Detects statistically significant triplet logic relationships from a binary matrix dataset (indicating connection, for instance co-occurrence, co-expression). Applies Logic Analysis of Phylogenetic Profiles (LAPP) method, which is based on normalized mutual information, to phylogenetic profiling data, but also applicable to gene co-expression and pathway data.	[37] https://github.com/fukunagatsu/LogicomeProfiler
CoCoNet	R package (via Github)	Integration of GWAS and gene-expression data	COmposite likelihood-based COvariance regression NETwork model to identify trait-relevant tissues or cell types. Uses covariance regression network models to express gene-level effect measurements for a given GWAS trait as a function of the tissue-specific co-expression adjacency matrix.	[38] http://www.xzlab.org/software.html
NEO	R package (via CRAN)	Integration of GWAS and gene-expression data	Network Edge Orienting infers directed gene networks by integrating gene-expression data with genetic marker data and compares them with structural equation models	[39] https://horvath.genetics.ucla.edu/html/aten/NEO/
WGCNA	R package (via CRAN)	Mainly gene-expression data, but can be applied to other omics	Weighted Gene Co-expression Network Analysis is used to find clusters, relating modules to one another and to external sample traits and calculates module membership measures. This approach facilitates gene screening and the identification of biomarkers.	[40,41] https://horvath.genetics.ucla.edu/html/CoexpressionNetwork/Rpackages/WGCNA/
DIABLO in mixOmics	R package (via Bioconductor)	All types (multi-omics)	Multivariate methods to analyse and visualize high-dimensional datasets (number of variables larger than number of samples). Complementary information from several data sets measured on the same N individuals, but across multiple omics data sets is combined to gain a better understanding of the interplay between the different levels of data that are measured (‘N-integration’). Data dimensions are reduced by applying sparse generalized canonical correlation analysis (SGCCA).	[42,43,44] http://mixomics.org/mixdiablo