Diagnostic Algorithm Application in Pediatric Patients with Complex Neurological Phenotypes in South Kazakhstan

Introduction: Complex neurological phenotypes (CNPs) in pediatric patients often involve multiple coexisting syndromes that challenge conventional diagnostic pathways. In the present study, we evaluated the diagnostic algorithm developed by our team—“Deep Phenotyping–Based Diagnostic Algorithm for Complex Neurological Phenotypes”—which integrates deep phenotyping and genomics.
Methods: In this single-center retrospective study, 28 children (0–18 years) with CNPs were assessed at the Diagnostic Center Clinic of Khoja Akhmet Yassawi International Kazakh-Turkish University between January and December 2023. Clinical features were encoded as standardized Human Phenotype Ontology (HPO) terms, which were entered into the Phenomizer platform to generate ranked candidate genes. Proband whole exome sequencing (WES) was performed for all participants. Reverse phenotyping re-examined clinical data in light of genetic findings. The algorithm’s diagnostic yield and concordance between prioritization and confirmed diagnoses were analyzed descriptively.
Results: A molecular diagnosis was established in 14 of 28 patients (50%). Confirmed etiologies included monogenic neurodevelopmental syndromes, metabolic disorders, mitochondrial encephalopathy, craniosynostosis, and chromosomal aberrations. In all single-gene cases, the causal gene ranked within Phenomizer’s top five candidates. Reverse phenotyping corroborated genotype–phenotype correlations and revealed additional clinical features in select patients.
Conclusions: This multi-step algorithm achieved a 50 % diagnostic yield—exceeding typical WES-only approaches—and demonstrated robust concordance between phenotype-driven prioritization and molecular results. The framework is feasible in resource-limited settings and may streamline the genetic workup of complex pediatric neurological disorders.