PhenoScore quantifies phenotypic variation for rare genetic diseases by combining facial analysis with other clinical features using a machine-learning framework.

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.

The phenotypic spectrum and genotype-phenotype correlations in 106 patients with variants in major autism gene CHD8.

CHD8, a major autism gene, functions in chromatin remodelling and has various roles involving several biological pathways. Therefore, unsurprisingly, previous studies have shown that intellectual developmental disorder with autism and macrocephaly (IDDAM), the syndrome caused by pathogenic variants in CHD8, consists of a broad range of phenotypic abnormalities. We collected and reviewed 106 individuals with IDDAM, including 36 individuals not previously published, thus enabling thorough genotype-phenotype analyses, involving the CHD8 mutation spectrum, characterization of the CHD8 DNA methylation episignature, and the systematic analysis of phenotypes collected in Human Phenotype Ontology (HPO). We identified 29 unique nonsense, 25 frameshift, 24 missense, and 12 splice site variants. Furthermore, two unique inframe deletions, one larger deletion (exons 26-28), and one translocation were observed. Methylation analysis was performed for 13 patients, 11 of which showed the previously established episignature for IDDAM (85%) associated with CHD8 haploinsufficiency, one analysis was inconclusive, and one showing a possible gain-of-function signature instead of the expected haploinsufficiency signature was observed. Consistent with previous studies, phenotypical abnormalities affected multiple organ systems. Many neurological abnormalities, like intellectual disability (68%) and hypotonia (29%) were observed, as well as a wide variety of behavioural abnormalities (88%). Most frequently observed behavioural problems included autism spectrum disorder (76%), short attention span (32%), abnormal social behaviour (31%), sleep disturbance (29%) and impaired social interactions (28%). Furthermore, abnormalities in the digestive (53%), musculoskeletal (79%) and genitourinary systems (18%) were noted. Although no significant difference in severity was observed between males and females, individuals with a missense variant were less severely affected. Our study provides an extensive review of all phenotypic abnormalities in patients with IDDAM and provides clinical recommendations, which will be of significant value to individuals with a pathogenic variant in CHD8, their families, and clinicians as it gives a more refined insight into the clinical and molecular spectrum of IDDAM, which is essential for accurate care and counselling.

Establishing the phenotypic spectrum of ZTTK syndrome by analysis of 52 individuals with variants in SON.

Zhu-Tokita-Takenouchi-Kim (ZTTK) syndrome, an intellectual disability syndrome first described in 2016, is caused by heterozygous loss-of-function variants in SON. Its encoded protein promotes pre-mRNA splicing of many genes essential for development. Whereas individual phenotypic traits have previously been linked to erroneous splicing of SON target genes, the phenotypic spectrum and the pathogenicity of missense variants have not been further evaluated. We present the phenotypic abnormalities in 52 individuals, including 17 individuals who have not been reported before. In total, loss-of-function variants were detected in 49 individuals (de novo in 47, inheritance unknown in 2), and in 3, a missense variant was observed (2 de novo, 1 inheritance unknown). Phenotypic abnormalities, systematically collected and analyzed in Human Phenotype Ontology, were found in all organ systems. Significant inter-individual phenotypic variability was observed, even in individuals with the same recurrent variant (n = 13). SON haploinsufficiency was previously shown to lead to downregulation of downstream genes, contributing to specific phenotypic features. Similar functional analysis for one missense variant, however, suggests a different mechanism than for heterozygous loss-of-function. Although small in numbers and while pathogenicity of these variants is not certain, these data allow for speculation whether de novo missense variants cause ZTTK syndrome via another mechanism, or a separate overlapping syndrome. In conclusion, heterozygous loss-of-function variants in SON define a recognizable syndrome, ZTTK, associated with a broad, severe phenotypic spectrum, characterized by a large inter-individual variability. These observations provide essential information for affected individuals, parents, and healthcare professionals to ensure appropriate clinical management.