Reliable machine learning models in genomic medicine using conformal prediction - perspective review

Aug 3, 2026

Upload by Miles

Machine learning and genomic medicine are central to personalized healthcare, enabling diagnosis, risk stratification, treatment tailoring, and adverse-effect prediction, yet prediction mistakes can carry life-threatening consequences in clinical practice. Conformal prediction is presented as a general framework to quantify predictive uncertainty, addressing reliability concerns. This perspective review surveys conformalized applications in genomic medicine, reviews key challenges for translating them to routine care, and evaluates both binary transductive and regression-based inductive drug-response modeling, alongside multi-class inductive handling of distribution shifts in molecular subtyping.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages14

Reading Time35 min

AccessFree to download

Introduction

Uncertainty quantification methods

UC Santa Cruz UC Santa Cruz Previously Published Works Title Reliable machine learning models in genomic medicine using conformal prediction. Permalink [https://escholarship.org/uc/item/37k7k4gs](https://escholarship.org/uc/item/37k7k4gs) Authors Papangelou, Christina Kyriakidis, Konstantinos Natsiavas, Panteliset al. Publication Date 2025 DOI 10.3389/fbinf.2025.1507448 Peer reviewed [eScholarship.org](eScholarship.org) Powered by the California Digital Library University of California TYPE Review PUBLISHED 24 February 2025 DOI 10.3389/fbinf.2025.1507448 OPEN ACCESS EDITED BY Natasa Miskov-Zivanov, University of Pittsburgh, United States REVIEWED BY David Alfredo Medina Ortiz, University of Magallanes, Chile Yannan Bin, Anhui University, China *CORRESPONDENCE Andigoni Malousi, [andigoni@auth.gr](andigoni@auth.gr) RECEIVED 07 October 2024 ACCEPTED 30 January 2025 PUBLISHED 24 February 2025 CITATION Papangelou C, Kyriakidis K, Natsiavas P, Chouvarda I and Malousi A (2025) Reliable machine learning models in genomic medicine using conformal prediction. Front. Bioinform. 5:1507448 . doi: 10.3389/fbinf.2025.1507448 COPYRIGHT © 2025 Papangelou, Kyriakidis, Natsiavas, Chouvarda and Malousi. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY) . The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms. Reliable machine learning models in genomic medicine using conformal prediction Christina Papangelou 1, Konstantinos Kyriakidis 2, Pantelis Natsiavas 3, Ioanna Chouvarda 2 and Andigoni Malousi 1,4* 1School of Medicine, Aristotle University of Thessaloniki, Thessaloniki, Greece, 2Genomics Institute, University of California Santa Cruz, Santa Cruz, CA, United States, 3 Institute of Applied Biosciences, Center for Research and Technology Hellas, Thessaloniki, Greece, 4GENeTres Research Group, Center for Interdisciplinary Research and Innovation, Thessaloniki, Greece Machine learning and genomic medicine are the mainstays of research in delivering personalized healthcare services for disease diagnosis, risk stratification, tailored treatment, and prediction of adverse effects. However, potential prediction errors in healthcare services can have life-threatening impact, raising reasonable skepticism about whether these applications have practical benefit in clinical settings. Conformal prediction offers a versatile framework for addressing these concerns by quantifying the uncertainty of predictive models. In this perspective review, we investigate potential applications of conformalized models in genomic medicine and discuss the challenges towards bridging genomic medicine applications with clinical practice. We also demonstrate the impact of a binary transductive model anda regression-based inductive model in predicting drug response as well asthe performance of a multi-class inductive predictor in addressing distribution shifts in molecular subtyping. The main conclusion is that as machine learning and genomic medicine are increasingly infiltrating healthcare services, conformal prediction has the potential to overcome the safety limitations of current methods and could be effectively integrated into uncertainty-informed applications within clinical environments. KEYWORDS conformal prediction, machine learning, genomic medicine, uncertainty estimate, reliable predictions, perspective review 1 Introduction Artificial Intelligence (AI)-based models are having transformative impact on highrisk predictions made for personalized medicine applications (Hamet and Tremblay, 2017) . Genomic medicine as a cornerstone of precision

FAQs

Why is reliability critical for machine learning in genomic medicine?

Prediction errors in healthcare can have life-threatening impact, making clinicians question whether ML-based genomic applications provide clear practical benefit in routine care.

How does conformal prediction address uncertainty in predictive models?

Conformal prediction quantifies the uncertainty of predictive models, helping move from single-point outputs toward uncertainty-informed predictions that support safer decision-making.

What modeling settings are discussed for drug response prediction?

The review demonstrates the impact of a binary transductive model and a regression-based inductive model for predicting drug response.

Aug 3, 2026

Upload by Miles