TABLE 4.
Low-level: Network-based unsupervised integration methods.
| Approach | Model | Macro category* | Author | Omics data** | Objective | Software*** |
|---|---|---|---|---|---|---|
| Matrix Factorization-based (MF-based) Networks | • CMF/CMF-W (Collective Matrix Factorization) | ModE | Liany et al. (2020) | Any Omics | Outcome/Interaction-prediction | • Python code (https://github.com/lianyh) |
| • NBS (Network-Based Stratification) | ModE | Hofree et al. (2013) | MiE, CNV, DM, GE, PE | Patient-subtyping | • pyNBS Python code (https://github.com/idekerlab/pyNBS) | |
| • DFMF (Data Fusion by Matrix Factorization) | ModE | Žitnik and Zupan, (2014) | GE, GO-terms, MeSH-descriptor | Gene function-prediction | • - | |
| • FUSENET | ModE | Žitnik and Zupan, (2015) | GE, Mutation | Disease-insight (Gene-Disease association- prediction) | • Python code (https://github.com/mims-harvard/fusenet) | |
| • Medusa | ModE | Zitnik and Zupan, (2016) | Any Omics | Module-discovery, Gene-Disease association- prediction | • Python code (https://github.com/mims-harvard/medusa) | |
| • MAE (Multi-view factorization AutoEncoder) | ModE | Ma and Zhang, (2019) | MiE, DM, GE, PE, PPIs | Disease-prediction | PyTorch code (https://github.com/BeautyOfWeb/Multiview-AutoEncoder) | |
| • DisoFun (Differentiate isoform Functions with collaborative matrix factorization) | ModE | Wang et al. (2020) | GE, IE | Disease-function Prediction | MATLAB code (http://mlda.swu.edu.cn/codes.php?%20name=DisoFun) | |
| • IMCDriver | DatE | Zhang et al. (2021) | GE, Mutation, PPIs | Gene-discovery | Python code (https://github.com/NWPU-903PR/IMCDriver) | |
| • RAIMC (RBP-AS Target Prediction Based on Inductive Matrix Completion) | ModE | Qiu et al. (2021) | AS, RBPs | Protein-prediction | MATLAB code (https://github.com/yushanqiu/RAIMC) | |
| Bayesian Networks (Pearl, 2014) (BNs) | • PARADIGM (PAthway Recognition Algorithm using Data Integration on Genomic Models) | ModE | Vaske et al. (2010) | CNV, GE, PE | Disease-subtyping, Disease-insight | • GIANT interface (http://giant.princeton.edu/) |
| • CONEXIC | ModE | Akavia et al. (2010) | GE, CNV | Gene-discovery | • - | |
| Network Propagation-based Networks (Random walk-, and Network Fusion-based Methods) | • GeneticInterPred | ModE | You et al. (2010) | GE, PE | Interaction-prediction | • - |
| • RWRM (Random Walk with Restart on Multigraphs) | ModE | Li and Li, (2012) | GE, PPIs | Gene-prioritizing | • - | |
| • TieDIE (Tied Diffusion through Interacting Events) | ModE | Paull et al. (2013) | GE, TF, PPIs | Module/sub-network detection | • Python code (https://sysbiowiki.soe.ucsc.edu/tiedie) | |
| • SNF (Similarity Network Fusion) | ModE | Wang et al. (2014) | MiE, DM, GE | Patient-subtyping | • SNFtool (https://cran.r-project.org/web/packages/SNFtool/index.html) | |
| • HotNet2 | ModE | Leiserson et al. (2015) | SNV, CNA, GE, PPIs | Sub-network detection | • HotNet software (http://compbio.cs.brown.edu/projects/hotnet/) | |
| • NetICS | ModE | Dimitrakopoulos et al. (2018) | MiE, CNV, GE | Biomarker-prediction | • Matlab code (https://github.com/cbg-ethz/netics) | |
| • RWR-M (Random Walk with Restart for Multiplex networks) | ModE | Valdeolivas et al. (2019) | GE, Co-expression, PPIs | Gene-prediction | • R code (https://github.com/alberto-valdeolivas/RWR-MH) | |
| • RWR-MH (RWR for Multiplex-Heterogeneous networks) | ModE | Valdeolivas et al. (2019) | GE, Co-expression, PPIs | Gene-prediction | • RandomWalkRestartMH (http://bioconductor.org/packages/release/bioc/html/RandomWalkRestartMH.html) | |
| • MSNE (Multiple Similarity Network Embedding) | ModE | Xu et al. (2020) | CNV, DM, GE | Disease-subtyping | • Python code (https://github.com/GaoLabXDU/MSNE) | |
| • RWRF (Random Walk with Restart for multi-dimensional data Fusion) | ModE | Wen et al. (2021) | MiE, DM, GE | Disease-subtyping | • R code (https://github.com/Sepstar/RWRF/) | |
| Correlation-based and Other Networks | • WGCNA (Weighted Gene Co-expression Network Analysis) | DatE | Langfelder and Horvath, (2008) | GE (from multiple platforms/species) | Gene-prioritizing | • WGCNA (https://horvath.genetics.ucla.edu/html/CoexpressionNetwork/Rpackages/WGCNA/) |
| • GGM (Gaussian Graphical Model) | ModE | Krumsiek et al. (2011) | SNP, GE, Met | Metabolite-pathway reactions | • - | |
| • GEM (GEnome scale Metabolic models) | ModE | Shoaie et al. (2013) | GE, Met | Metabolite-subnetwork | • - | |
| • DBN (Deep Belief Network) | ModE | Liang et al. (2014) | MiE, DM, GE | Disease-subtyping | • Python code (https://github.com/glgerard/MDBN) | |
| • Lemon-Tree | ModE | Bonnet et al. (2015) | CNV, GE | Biomarker-discovery | • JAVA command (https://github.com/erbon7/lemon-tree) | |
| • TransNet (Transkingdom Network) | ModE | Rodrigues et al. (2018) | Any Omics | Causal network | • TransNetDemo R code (https://github.com/richrr/TransNetDemo) |
*Main categories include (A) Multi-step and Sequential Analysis (MS-SA), (B) Data-ensemble (DatE), (C) Model-ensemble (ModE). ** CNV: copy number variation, CAN: copy number alternation, SNV: single nucleotide variation, DM: DNA methylation, AS: alternative splicing, MiE: Micro RNA expression, GE: gene expression, TF: transcriptional factor, IE: isoform expression, PE: protein expression, RBPs: RNA-Binding Proteins, PPI: Protein-protein interactions, Met: Metabolite. ***R packages, unless otherwise stated.