Dealing with dimensionality: the application of machine learning to multi-omics data

Abstract

Motivation: Machine learning (ML) methods are motivated by the need to automate information extraction from large datasets in order to support human users in data-driven tasks. This is an attractive approach for integrative joint analysis of vast amounts of omics data produced in next generation sequencing and other -omics assays. A systematic assessment of the current literature can help to identify key trends and potential gaps in methodology and applications. We surveyed the literature on ML multi-omic data integration and quantitatively explored the goals, techniques and data involved in this field. We were particularly interested in examining how researchers use ML to deal with the volume and complexity of these datasets.

Results: Our main finding is that the methods used are those that address the challenges of datasets with few samples and many features. Dimensionality reduction methods are used to reduce the feature count alongside models that can also appropriately handle relatively few samples. Popular techniques include autoencoders, random forests and support vector machines. We also found that the field is heavily influenced by the use of The Cancer Genome Atlas dataset, which is accessible and contains many diverse experiments.

Availability and implementation: All data and processing scripts are available at this GitLab repository: https://gitlab.com/polavieja_lab/ml_multi-omics_review/ or in Zenodo: https://doi.org/10.5281/zenodo.7361807.

Supplementary information: Supplementary data are available at Bioinformatics online.

Fig. 1.

Overview of paper selection process

Fig. 2.

(a) Number of uses of each -omics category in the reviewed papers. (b) Number of -omics used per paper in the reviewed papers. (c) The number of appearances -omics pairs across papers

Fig. 3.

‘Shape’ of multi-omics datasets. Number of samples (x-axis), number of features (y-axis)

Fig. 4.

(a) Number of ML techniques being used more than once in the reviewed papers. The publications on ‘machine learning AND multi-omics AND integration’ are plotted in green, while the publications on ‘machine learning AND integration’ are plotted in purple. Significant differences were observed for autoencoders, and Cox proportional hazards (Cox PH) (Cox, 1972). Fisher’s exact test P-values of <0.0001 in both cases, satisfying the Bonferroni correction for this number of tests. Number in the reviewed multi-omics ML papers of ML goals (b) and labels used for classification (c)

Fig. 5.

Number of citations per year, of papers published in different years

Fig. 6.

Shapes in datasets. (a) A dataset where n ≫ P, the ideal ‘shape’ for many ML techniques. (b) In multi-omics analyses, researchers face very wide datasets, where n ≪ P. (c) Feature selection and extraction are often used to reduce the number of features. In feature selection, a subset of the original features is kept. In feature extraction, features are merged and transformed into a smaller number of new ones