Personalised Drug Identifier for Cancer Treatment with Transformers using Auxiliary Information

TL;DR

This paper introduces a transformer-based model for personalized drug response prediction in cancer treatment, leveraging auxiliary patient data and addressing variable mutation sequences, leading to improved accuracy and real-world clinical deployment.

Contribution

It presents a novel transformer approach that explicitly models mutation sequences and incorporates auxiliary information, outperforming existing models in drug response prediction.

Findings

Surpasses state-of-the-art DRP models on benchmark data.

Deployed a treatment recommendation system in a clinical setting.

Achieves improved prediction accuracy using auxiliary patient data.

Abstract

Cancer remains a global challenge due to its growing clinical and economic burden. Its uniquely personal manifestation, which makes treatment difficult, has fuelled the quest for personalized treatment strategies. Thus, genomic profiling is increasingly becoming part of clinical diagnostic panels. Effective use of such panels requires accurate drug response prediction (DRP) models, which are challenging to build due to limited labelled patient data. Previous methods to address this problem have used various forms of transfer learning. However, they do not explicitly model the variable length sequential structure of the list of mutations in such diagnostic panels. Further, they do not utilize auxiliary information (like patient survival) for model training. We address these limitations through a novel transformer based method, which surpasses the performance of state-of-the-art DRP models on benchmark data. We also present the design of a treatment recommendation system (TRS), which is currently deployed at the National University Hospital, Singapore and is being evaluated in a clinical trial.