A Federated Learning Benchmark for Drug-Target Interaction

02/15/2023

∙

Aggregating pharmaceutical data in the drug-target interaction (DTI) domain has the potential to deliver life-saving breakthroughs. It is, however, notoriously difficult due to regulatory constraints and commercial interests. This work proposes the application of federated learning, which we argue to be reconcilable with the industry's constraints, as it does not require sharing of any information that would reveal the entities' data or any other high-level summary of it. When used on a representative GraphDTA model and the KIBA dataset it achieves up to 15 available non-privacy preserving alternative. Our extensive battery of experiments shows that, unlike in other domains, the non-IID data distribution in the DTI datasets does not deteriorate FL performance. Additionally, we identify a material trade-off between the benefits of adding new data, and the cost of adding more clients.

READ FULL TEXT

A Federated Learning Benchmark for Drug-Target Interaction

FedMix: Approximation of Mixup under Mean Augmented Federated Learning

Clustered Federated Learning: Model-Agnostic Distributed Multi-Task Optimization under Privacy Constraints

Efficient Federated Learning on Knowledge Graphs via Privacy-preserving Relation Embedding Aggregation

Towards Fair and Privacy Preserving Federated Learning for the Healthcare Domain

Federated Learning with Heterogeneous Architectures using Graph HyperNetworks

A Federated Learning Benchmark for Drug-Target Interaction

Related Research

FedMix: Approximation of Mixup under Mean Augmented Federated Learning

Clustered Federated Learning: Model-Agnostic Distributed Multi-Task Optimization under Privacy Constraints

Efficient Federated Learning on Knowledge Graphs via Privacy-preserving Relation Embedding Aggregation

Towards Fair and Privacy Preserving Federated Learning for the Healthcare Domain

Federated Learning with Heterogeneous Architectures using Graph HyperNetworks