Back

Topology Adaptive Graph Convolution Network With Heterogeneous Entities For Predicting Adverse Events from Drug-Drug-Interactions

Bagchi, S.; Sarkar, A.; Maulik, U.

2022-05-17 bioinformatics
10.1101/2022.05.16.491112 bioRxiv
Show abstract

In times like this, it is imperative to be cautious about the effects of drugs or vaccination doses on patients who are already suffering from other serious diseases. Its not only the virus which can affect the body metabolisms, drugs to encounter the virus may also end up having unwanted negative effects. Therapeutic activities of drugs are often influenced by co-administration of drugs that may cause inevitable drug-drug interactions and inadvertent side effects. Therefore, prediction and identification of DDIs are extremely vital for patient safety and treatment modalities. In this context, we intend to develop a computational method based on functional similarity of drugs. Our objective has been enhanced by the usage of knowledge graphs which allows capturing underlying information embeddings in a biological network with heterogenous entities. On providing this knowledge graph as the input to a Topology Adaptive Graph Convolution Network which performs topologically-aware flexible convolutions, we achieve improvements on priorly proposed GCN models that has been shown as comparison.

Matching journals

The top 8 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.