Background

Pediatric tumor boards increasingly receive multi-omic data per patient — whole-exome and RNA sequencing, methylation profiling, and (when available) organoid drug-response screens. The volume and heterogeneity of this data outpace what any single reviewer can synthesise consistently across cases. Decisions still come down to expert judgement, and rightly so, but the field needs tools that make the underlying evidence legible and consistent across institutions.

Objective

PEDICTOR is a decision-support algorithm that ingests per-patient multi-omic and (where available) functional drug-screen data and produces a ranked, evidence-linked therapeutic recommendation set for tumor-board discussion. It is explicitly designed as a support tool — not a replacement for the board — and its outputs are framed to make sources, confidence, and dissenting signals visible.

Aims

• Integrate sequencing, methylation, and organoid drug-screen data with curated pediatric oncology evidence (PMTL, pediatric precision-medicine literature, ongoing trials).
• Quantify concordance with expert consensus tumor-board recommendations and characterise the discordant cases.
• Prospectively validate the tool at three pediatric cancer centers under an IRB-approved observational study, with patient outcomes followed for at least 24 months.

The program is supported by an institutional frontier-program award. PEDICTOR is co-developed with a university bioengineering research group and is intended for non-commercial academic use; a deployable open-source build will follow successful prospective validation.