Fig. 4

Integrative strategies of multi-omics data in machine learning. Machine learning methods for integration of multi-omics data are mainly categorized into three types. The first type of integrative strategy combines directly each omics data into one large matrix before training machine learning model. The second type of integrative strategy firstly transforms each omics data into an intermediate form that then are combined for further training. The third type of integrative strategy firstly trains machine learning model separately using each omics data and then combines the predictive outcomes of the trained individual models to generate a finalized model