Both automatic and not automatic MOFA analysis includes three methods to ease the user interpretation of the discriminating MOFA factors.
Correlation analysis: Users can upload a matrix containing binary or numerical clinical features to integrate into the MOFA model. The numerical data are correlated through a Pearson correlation. The binary clinical data are analysed using the Wilcoxon test after dividing the groups into positive or negative categories. The nominal p.values were corrected using the Benjamini-Hochberg method.
Pathway analysis: BiomiX retrieves the top contributing genes, metabolites and CpG island for discriminating factors in each MOFA model. Depending on the type of omics data is selected an R package to highlight if there is an enrichment of some biological or metabolic pathway within the enriched genes, metabolites and CpG island. The genes are analysed by EnrichR using the "Reactome", "GO biological process" and "CODE and ChEA consensus TFs from ChIP-X" libraries, while the metabolites are analysed through MetPath using the KEGG and HMDB databases. The CpG islands are associated with their genes, if it exists and is examined through EnrichR.
Pubmed bibliography research: For each discriminating factor in each MOFA model, the top contributing genes, metabolites and CpG island genes are used as input in Pubmed research. The aim is to retrieve articles' abstracts associated with each discriminating factor, to have each of them clues behind their identity. The search algorithm has three levels of research, where the results merging more multiomics contributors are selected. At first, the algorithm selects the top 10 contributors from each omic provided as input and carries out research in which only the article abstracts show at least one out of ten contributors for each omic in the text. The second level does the same, but it selects article abstracts containing at least one out of ten contributors in omics pairs (e.g. transcriptomics-metabolomics and methylomics-transcriptomics, metabolomics-methylomics). Finally, the last-level research article abstracts show at least one out of ten contributors within a single omics in the text. The output document includes, for each of these three levels, an Excel document including the PubMed articles, the total number of matches with the contributors and how many contributors are matched in the text. Furthermore, keywords, DOI and the match information for each contributor are also available. The authors-provided keywords are not optimal due to their absence in some journals. That is why BiomiX includes here a further text mining analysis. The article's abstract spotted in each level is extracted and parsed through the R lit search v 1.0.0 in a combination of 2-4 words. All the vocabularies generated by each abstract are analysed to spot combinations of words more frequently. These words are filtered by another vocabulary made up from gene set names as GO biological process (7751 gene sets) and Human Phenotype Ontology (5405 gene sets). The most frequent 15 words are included in the output Excel file. To conclude, the same analysis is repeated again, including all the abstracts retrieved within the three-level.