PhenoScore Quantifies Phenotypic Variation for Rare Genetic Diseases by Combining Facial Analysis With Other Clinical Features Using a Machine-Learning Framework

Authors

Alexander J M Dingemans
Max Hinne
Kim M G Truijen
Lia Goltstein
Jeroen van Reeuwijk
Nicole de Leeuw
Janneke Schuurs-Hoeijmakers
Rolph Pfundt
Illja J Diets
Joery den Hoed
Elke de Boer
Jet Coenen-van der Spek
Sandra Jansen
Bregje W van Bon
Noraly Jonis
Charlotte W Ockeloen
Anneke T Vulto-van Silfhout
Tjitske Kleefstra
David A Koolen
Philippe M Campeau
Elizabeth E Palmer
Hilde Van Esch
Gholson J Lyon
Fowzan S Alkuraya
Anita Rauch
Ronit Marom
Diana Baralle
Pleuntje J van der Sluijs
Gijs W E Santen
R Frank Kooy
Marcel A J van Gerven
Lisenka E L M Vissers
Bert B A de Vries

Publication Date

9-1-2023

Journal

Nature Genetics

DOI

10.1038/s41588-023-01469-w

PMID

37550531

PMCID

PMC11414844

PubMedCentral® Posted Date

9-20-2024

PubMedCentral® Full Text Version

Author MSS

Published Open-Access

yes

Keywords

Humans, Artificial Intelligence, Phenotype, Algorithms, Machine Learning, Biological Variation, Population, Matrix Attachment Region Binding Proteins, DNA-Binding Proteins, Transcription Factors, variant of unknown significance, rare disease, personalized medicine, facial recognition, deep phenotyping

Abstract

Several molecular and phenotypic algorithms exist that establish genotype-phenotype correlations, including facial recognition tools. However, no unified framework that investigates both facial data and other phenotypic data directly from individuals exists. We developed PhenoScore: an open-source, artificial intelligence-based phenomics framework, combining facial recognition technology with Human Phenotype Ontology data analysis to quantify phenotypic similarity. Here we show PhenoScore's ability to recognize distinct phenotypic entities by establishing recognizable phenotypes for 37 of 40 investigated syndromes against clinical features observed in individuals with other neurodevelopmental disorders and show it is an improvement on existing approaches. PhenoScore provides predictions for individuals with variants of unknown significance and enables sophisticated genotype-phenotype studies by testing hypotheses on possible phenotypic (sub)groups. PhenoScore confirmed previously known phenotypic subgroups caused by variants in the same gene for SATB1, SETBP1 and DEAF1 and provides objective clinical evidence for two distinct ADNP-related phenotypes, already established functionally.