Concatenation-based integration
Concatenation-based integration combines multiple omic datasets, raw or pre-processed, into a single large matrix. One of the advantages of these approaches is their simplicity, since once the concatenation of multi-omic datasets is achieved, unsupervised and supervised analysis methods can be applied to the joint matrix, as in the case of the independent analysis of omic layers. Concatenation-based techniques offer a straightforward approach to utilising machine learning for the examination of both continuous and categorical data. Once the individual omics are concatenated, these methods can analyse all the combined features in an even-handed manner and pinpoint the most distinguishing features associated with a given phenotype. One of the main challenges of concatenation-based approaches is to ensure that the features of the different omic layers are comparable.
Several examples of unsupervised concatenation-based methods for multi-omic integration have been developed in recent years, most of them based on matrix-factorisation [58]. Joint non-negative matrix factorisation (Joint NMF) allowed integrating non-negative multi-omic data by decomposing the joint matrix into factors and loadings [28]. Joint and Individual Variation Explained (JIVE) is an adaptation of NMF framework [59] which was later improved by Joint Bayes Factor (JBF) to handle the problems derived from the high sparsity of multi-omic datasets [60]. iCluster framework is based in similar principles to NMF but allows integration of datasets having negative values [61]. MoCluster [62], RLAcluster [63] and iClusterBayes [64] have further developed the framework and improved it in terms of diversity of handled data types, computation speed and clustering accuracy. Multi-Omics Factor Analysis (MOFA) is another recent development that allows discovering the principal sources of variability across different omic datasets [65]. Regarding supervised analyses, any of the algorithms for supervised analysis of single omic layers can be used to analyse concatenated multi omic data. RF [66], SVM [67], LASSO regression [68] or DL [69] algorithms have been used, among others, for concatenation-based supervised analysis in multi-omic literature.
28. Zhang B, Brock M, Arana C, Dende C, Oers NS van, Hooper LV, et al. Impact of Bead-Beating intensity on the genus- and Species-Level characterization of the gut microbiome using amplicon and complete 16S rRNA gene sequencing. Front Cell Infect Microbiol. 2021;11:678522.
58. Reel PS, Reel S, Pearson E, Trucco E, Jefferson E. Using machine learning approaches for multi-omics data analysis: A review. Biotechnol Adv. 2021;49:107739.
59. Lock EF, Hoadley KA, Marron JS, Nobel AB. JOINT AND INDIVIDUAL VARIATION EXPLAINED (JIVE) FOR INTEGRATED ANALYSIS OF MULTIPLE DATA TYPES. Ann Appl Stat. 2013;7:523–42.
60. Ray P, Zheng L, Lucas J, Carin L. Bayesian joint analysis of heterogeneous genomics data. Bioinformatics. 2014;30:1370–6.
61. Shen R, Olshen AB, Ladanyi M. Integrative clustering of multiple genomic data types using a joint latent variable model with application to breast and lung cancer subtype analysis. Bioinformatics. 2009;25:2906–12.
62. Meng C, Helm D, Frejno M, Kuster B. moCluster: Identifying joint patterns across multiple omics data sets. J Proteome Res. 2016;15:755–65.
63. Wu D, Wang D, Zhang MQ, Gu J. Fast dimension reduction and integrative clustering of multi-omics data using low-rank approximation: Application to cancer molecular classification. BMC Genomics. 2015;16:1022.
64. Mo Q, Shen R, Guo C, Vannucci M, Chan KS, Hilsenbeck SG. A fully bayesian latent variable model for integrative clustering analysis of multi-type omics data. Biostatistics. 2018;19:71–86.
65. Argelaguet R, Velten B, Arnol D, Dietrich S, Zenz T, Marioni JC, et al. Multi-Omics factor analysis—a framework for unsupervised integration of multi-omics data sets. Mol Syst Biol. 2018;14:e8124.
66. Acharjee A, Kloosterman B, Visser RGF, Maliepaard C. Integration of multi-omics data for prediction of phenotypic traits using random forest. BMC Bioinformatics. 2016;17 Suppl 5 Suppl 5:180.
67. Li S, Chen X, Liu X, Yu Y, Pan H, Haak R, et al. Complex integrated analysis of lncRNAs-miRNAs-mRNAs in oral squamous cell carcinoma. Oral Oncol. 2017;73:1–9.
68. Lee G, Bang L, Kim SY, Kim D, Sohn K-A. Identifying subtype-specific associations between gene expression and DNA methylation profiles in breast cancer. BMC Med Genomics. 2017;10 Suppl 1:28.
69. Zhang L, Lv C, Jin Y, Cheng G, Fu Y, Yuan D, et al. Deep Learning-Based Multi-Omics data integration reveals two prognostic subtypes in High-Risk neuroblastoma. Front Genet. 2018;9:477.