Drug-drug interactions (DDIs) can occur when two or more drugs are administered, leading to side effects beyond those observed when drugs are taken by themselves. As the number of possible drug pairs is enormous, it is nearly impossible to experimentally test all combinations and discover previously unobserved side effects. This is where machine learning can help.
This study proposes a Siamese self-attention multi-modal neural network for DDI prediction. AttentionDDI integrates multiple drug similarity measures derived from a comparison of drug characteristics, including drug targets, pathways and gene expression profiles. The proposed model is able to accurately predict DDIs and beneficially applies an Attention mechanism, typically used in the Natural Language Processing domain, to aid in DDI model explainability.