English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Precision oncology, the genetic sequencing of tumors to identify druggabletargets, has emerged as the standard of care in the treatment of many cancers.Nonetheless, due to the pace of therapy development and variability in patientinformation, designing effective protocols for individual treatment assignmentin a sample-efficient way remains a major challenge. One promising approach tothis problem is to frame precision oncology treatment as a contextual banditproblem and to apply sequential decision-making algorithms designed to minimizeregret in this setting. However, a clear prerequisite for considering thismethodology in high-stakes clinical decisions is careful benchmarking tounderstand realistic costs and benefits. Here, we propose a benchmark datasetto evaluate contextual bandit algorithms based on real in vitro drug responseof approximately 900 cancer cell lines. Specifically, we curated a dataset ofcomplete treatment responses for a subset of 7 treatments from prior in vitrostudies. This allows us to compute the regret of proposed decision policiesusing biologically plausible counterfactuals. We ran a suite of Bayesian banditalgorithms on our benchmark, and found that the methods accumulate less regretover a sequence of treatment assignment tasks than a rule-based baselinederived from current clinical practice. This effect was more pronounced whengenomic information was included as context. We expect this work to be astarting point for evaluation of both the unique structural requirements andethical implications for real-world testing of bandit based clinical decisionsupport.

A Biologically Plausible Benchmark for Contextual Bandit Algorithms in Precision Oncology Using in vitro Data