| Popis: |
IntroductionGenomic conditions can be associated with developmental delay, intellectual disability and physical and mental health symptoms, but are individually rare and variable, which limits the use of standard clinical guidelines. A simple screening tool to identify young people with genetic conditions associated with neurodevelopmental disorders (ND-GC) who could benefit from further support would be of considerable value. We used machine learn approaches to address this question.MethodsA total of 489 individuals were included: 376 with a ND-GC, mean age=9.33, 63% male) and 113 unaffected siblings; mean age=10.35, 50% male). Primary carers completed detailed assessments of behavioural, neurodevelopmental and psychiatric symptoms and physical health conditions. Machine learning techniques (elastic net regression, random forests, support vector machines and artificial neural networks) were used to develop classifiers of ND-GC status using a limited set of variables. Exploratory Graph Analysis was used to understand associations within the final variable set.ResultsWe identified a set of 30 variables best discriminating between ND-GC carriers and control individuals, which formed 4 dimensions: Anxiety, Motor Development, Insomnia and Depression. All methods showed high discrimination accuracy with Linear Support Vector machines outperforming other methods (AUROC between 0.959 and 0.971).ConclusionsIn this study we developed models that identified a compact set of psychiatric and physical health measures that differentiate individuals with a ND-GC from controls and highlight the structure within these measures. This work is a step toward developing of a screening instrument to select young people with ND-GCs who might benefit from further specialist assessment. |
