A major controversy in psychiatric genetics is whether nonadditive genetic interaction effects contribute to the risk of highly polygenic disorders. We applied a support vector machines (SVMs) ...approach, which is capable of building linear and nonlinear models using kernel methods, to classify cases from controls in a large schizophrenia case–control sample of 11,853 subjects (5,554 cases and 6,299 controls) and compared its prediction accuracy with the polygenic risk score (PRS) approach. We also investigated whether SVMs are a suitable approach to detecting nonlinear genetic effects, that is, interactions. We found that PRS provided more accurate case/control classification than either linear or nonlinear SVMs, and give a tentative explanation why PRS outperforms both multivariate regression and linear kernel SVMs. In addition, we observe that nonlinear kernel SVMs showed higher classification accuracy than linear SVMs when a large number of SNPs are entered into the model. We conclude that SVMs are a potential tool for assessing the presence of interactions, prior to searching for them explicitly.
Rare copy number variants (CNVs) are associated with risk of neurodevelopmental disorders characterised by varying degrees of cognitive impairment, including schizophrenia, autism spectrum disorder ...and intellectual disability. However, the effects of many individual CNVs in carriers without neurodevelopmental disorders are not yet fully understood, and little is known about the effects of reciprocal copy number changes of known pathogenic loci.AimsWe aimed to analyse the effect of CNV carrier status on cognitive performance and measures of occupational and social outcomes in unaffected individuals from the UK Biobank.
We called CNVs in the full UK Biobank sample and analysed data from 420 247 individuals who passed CNV quality control, reported White British or Irish ancestry and were not diagnosed with neurodevelopmental disorders. We analysed 33 pathogenic CNVs, including their reciprocal deletions/duplications, for association with seven cognitive tests and four general measures of functioning: academic qualifications, occupation, household income and Townsend Deprivation Index.
Most CNVs (24 out of 33) were associated with reduced performance on at least one cognitive test or measure of functioning. The changes on the cognitive tests were modest (average reduction of 0.13 s.d.) but varied markedly between CNVs. All 12 schizophrenia-associated CNVs were associated with significant impairments on measures of functioning.
CNVs implicated in neurodevelopmental disorders, including schizophrenia, are associated with cognitive deficits, even among unaffected individuals. These deficits may be subtle but CNV carriers have significant disadvantages in educational attainment and ability to earn income in adult life.Declaration of interestNone.
Genomic CNVs increase the risk for early-onset neurodevelopmental disorders, but their impact on medical outcomes in later life is still poorly understood. The UK Biobank allows us to study the ...medical consequences of CNVs in middle and old age in half a million well-phenotyped adults.
We analysed all Biobank participants for the presence of 54 CNVs associated with genomic disorders or clinical phenotypes, including their reciprocal deletions or duplications. After array quality control and exclusion of first-degree relatives, we compared 381 452 participants of white British or Irish origin who carried no CNVs with carriers of each of the 54 CNVs (ranging from 5 to 2843 persons). We used logistic regression analysis to estimate the risk of developing 58 common medical phenotypes (3132 comparisons).
Many of the CNVs have profound effects on medical health and mortality, even in people who have largely escaped early neurodevelopmental outcomes. Forty-six CNV-phenotype associations were significant at a false discovery rate threshold of 0.1, all in the direction of increased risk. Known medical consequences of CNVs were confirmed, but most identified associations are novel. Deletions at 16p11.2 and 16p12.1 had the largest numbers of significantly associated phenotypes (seven each). Diabetes, hypertension, obesity and renal failure were affected by the highest numbers of CNVs. Our work should inform clinicians in planning and managing the medical care of CNV carriers.
Machine learning methods have been employed to make predictions in psychiatry from genotypes, with the potential to bring improved prediction of outcomes in psychiatric genetics; however, their ...current performance is unclear. We aim to systematically review machine learning methods for predicting psychiatric disorders from genetics alone and evaluate their discrimination, bias and implementation. Medline, PsycInfo, Web of Science and Scopus were searched for terms relating to genetics, psychiatric disorders and machine learning, including neural networks, random forests, support vector machines and boosting, on 10 September 2019. Following PRISMA guidelines, articles were screened for inclusion independently by two authors, extracted, and assessed for risk of bias. Overall, 63 full texts were assessed from a pool of 652 abstracts. Data were extracted for 77 models of schizophrenia, bipolar, autism or anorexia across 13 studies. Performance of machine learning methods was highly varied (0.48-0.95 AUC) and differed between schizophrenia (0.54-0.95 AUC), bipolar (0.48-0.65 AUC), autism (0.52-0.81 AUC) and anorexia (0.62-0.69 AUC). This is likely due to the high risk of bias identified in the study designs and analysis for reported results. Choices for predictor selection, hyperparameter search and validation methodology, and viewing of the test set during training were common causes of high risk of bias in analysis. Key steps in model development and validation were frequently not performed or unreported. Comparison of discrimination across studies was constrained by heterogeneity of predictors, outcome and measurement, in addition to sample overlap within and across studies. Given widespread high risk of bias and the small number of studies identified, it is important to ensure established analysis methods are adopted. We emphasise best practices in methodology and reporting for improving future studies.
Levels of activity are often affected in psychiatric disorders and can be core symptoms of illness. Advances in technology now allow the accurate assessment of activity levels but it remains unclear ...whether alterations in activity arise from shared risk factors for developing psychiatric disorders, such as genetics, or are better explained as consequences of the disorders and their associated factors. We aimed to examine objectively-measured physical activity in individuals with psychiatric disorders, and assess the role of genetic liability for psychiatric disorders on physical activity. Accelerometer data were available on 95,529 UK Biobank participants, including measures of overall mean activity and minutes per day of moderate activity, walking, sedentary activity, and sleep. Linear regressions measured associations between psychiatric diagnosis and activity levels, and polygenic risk scores (PRS) for psychiatric disorders and activity levels. Genetic correlations were calculated between psychiatric disorders and different types of activity. Having a diagnosis of schizophrenia, bipolar disorder, depression, or autism spectrum disorders (ASD) was associated with reduced overall activity compared to unaffected controls. In individuals without a psychiatric disorder, reduced overall activity levels were associated with PRS for schizophrenia, depression, and ASD. ADHD PRS was associated with increased overall activity. Genetic correlations were consistent with PRS findings. Variation in physical activity is an important feature across psychiatric disorders. Whilst levels of activity are associated with genetic liability to psychiatric disorders to a very limited extent, the substantial differences in activity levels in those with psychiatric disorders most likely arise as a consequences of disorder-related factors.
Celotno besedilo
Dostopno za:
DOBA, IZUM, KILJ, NUK, PILJ, PNG, SAZU, SIK, UILJ, UKNU, UL, UM, UPUK
Abstract
Alzheimer’s disease is a neurodegenerative disorder and the most common form of dementia. Early diagnosis may assist interventions to delay onset and reduce the progression rate of the ...disease. We systematically reviewed the use of machine learning algorithms for predicting Alzheimer’s disease using single nucleotide polymorphisms and instances where these were combined with other types of data. We evaluated the ability of machine learning models to distinguish between controls and cases, while also assessing their implementation and potential biases. Articles published between December 2009 and June 2020 were collected using Scopus, PubMed and Google Scholar. These were systematically screened for inclusion leading to a final set of 12 publications. Eighty-five per cent of the included studies used the Alzheimer's Disease Neuroimaging Initiative dataset. In studies which reported area under the curve, discrimination varied (0.49–0.97). However, more than half of the included manuscripts used other forms of measurement, such as accuracy, sensitivity and specificity. Model calibration statistics were also found to be reported inconsistently across all studies. The most frequent limitation in the assessed studies was sample size, with the total number of participants often numbering less than a thousand, whilst the number of predictors usually ran into the many thousands. In addition, key steps in model implementation and validation were often not performed or unreported, making it difficult to assess the capability of machine learning models.
This review assessed the literature regarding the performance of machine learning for the prediction of Alzheimer’s disease using genetics. Publications were screened systematically, with 12 studies passing all criteria. Common issues such as the overreliance on one data source, insufficient sample sizes and inconsistent model performance reporting methods were outlined.
Graphical Abstract
Graphical Abstract
Abstract Mutations in the LRRK2 gene are the most common genetic cause of familial Parkinson’s Disease (LRRK2-PD) and an important risk factor for sporadic PD (sPD). Multiple clinical trials are ...ongoing to evaluate the benefits associated with the therapeutical reduction of LRRK2 kinase activity. In this study, we described the changes of transcriptomic profiles (whole blood mRNA levels) of LRRK2 protein interactors in sPD and LRRK2-PD cases as compared to healthy controls with the aim of comparing the two PD conditions. We went on to model the protein-protein interaction (PPI) network centred on LRRK2, which was weighted to reflect the transcriptomic changes on expression and co-expression levels of LRRK2 protein interactors. Our results showed that LRRK2 interactors present both similar and distinct alterations in expression levels and co-expression behaviours in the sPD and LRRK2-PD cases; suggesting that, albeit being classified as the same disease based on clinical features, LRRK2-PD and sPD display significant differences from a molecular perspective. Interestingly, the similar changes across the two PD conditions result in decreased connectivity within a topological cluster of the LRRK2 PPI network associated with protein metabolism/biosynthesis and ribosomal metabolism suggesting protein homoeostasis and ribosomal dynamics might be affected in both sporadic and familial PD in comparison with controls.
Genomic conditions can be associated with developmental delay, intellectual disability, autism spectrum disorder, and physical and mental health symptoms. They are individually rare and highly ...variable in presentation, which limits the use of standard clinical guidelines for diagnosis and treatment. A simple screening tool to identify young people with genomic conditions associated with neurodevelopmental disorders (ND-GCs) who could benefit from further support would be of considerable value. We used machine learning approaches to address this question.
A total of 493 individuals were included: 389 with a ND-GC, mean age = 9.01, 66% male) and 104 siblings without known genomic conditions (controls, mean age = 10.23, 53% male). Primary carers completed assessments of behavioural, neurodevelopmental and psychiatric symptoms and physical health and development. Machine learning techniques (penalised logistic regression, random forests, support vector machines and artificial neural networks) were used to develop classifiers of ND-GC status and identified limited sets of variables that gave the best classification performance. Exploratory graph analysis was used to understand associations within the final variable set.
All machine learning methods identified variable sets giving high classification accuracy (AUROC between 0.883 and 0.915). We identified a subset of 30 variables best discriminating between individuals with ND-GCs and controls which formed 5 dimensions: conduct, separation anxiety, situational anxiety, communication and motor development.
This study used cross-sectional data from a cohort study which was imbalanced with respect to ND-GC status. Our model requires validation in independent datasets and with longitudinal follow-up data for validation before clinical application.
In 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 higher-order structure within these measures. This work is a step towards developing a screening instrument to identify young people with ND-GCs who might benefit from further specialist assessment.
IMPORTANCE: The role of large, rare copy number variants (CNVs) in neuropsychiatric disorders is well established, but their association with common psychiatric disorders, such as depression, remains ...unclear. OBJECTIVE: To examine the association of a group of 53 CNVs associated with neurodevelopmental disorders and burden of rare CNVs with risk of depression. DESIGN, SETTING, AND PARTICIPANTS: This case-control study used data from the UK Biobank study sample, which comprised 502 534 individuals living in the United Kingdom. Individuals with autism spectrum disorder, intellectual disability, attention-deficit/hyperactivity disorder, schizophrenia, or bipolar affective disorder diagnoses were excluded. Analyses were further restricted to individuals of European genetic ancestry (n = 407 074). The study was conducted from January 2017 to September 2018. EXPOSURES: CNV carrier status. MAIN OUTCOMES AND MEASURES: For the primary outcome, individuals who reported that a physician had told them they had a depression diagnosis were defined as cases. Analyses were repeated using 2 alternative depression definitions: self-reported lifetime depression with current antidepressant prescription at the time of visit 1, and hospital discharge diagnosis of depression. RESULTS: Copy number variants were identified in 488 366 individuals aged 37 to 73 years. In total, 407 074 individuals with European genetic ancestry (220 201 female 54.1%; mean SD age of 56.9 8.0 years) were included in the study. Of these individuals, 23 979 (5.9%) had self-reported lifetime depression and 383 095 (94.1%) reported no lifetime depression. The group of 53 neurodevelopmental CNVs was associated with self-reported depression (odds ratio OR, 1.34; 95% CI, 1.19-1.49, uncorrected P = 1.38 × 10−7), and these results were consistent when using 2 alternative definitions of depression. This association was partially explained by physical health, educational attainment, social deprivation, smoking status, and alcohol consumption. A strong independent association remained between the neurodevelopmental CNVs and depression in analyses that incorporated these other measures (OR, 1.26; 95% CI, 1.11-1.43; P = 2.87 × 10−4). Eight individual CNVs were nominally associated with risk of depression, and 3 of these 8 CNVs (1q21.1 duplication, Prader-Willi syndrome duplication, and 16p11.2 duplication) survived Bonferroni correction for the 53 CNVs tested. After the exclusion of carriers of neurodevelopmental CNVs, no association was found between measures of CNV burden and depression. CONCLUSIONS AND RELEVANCE: Neurodevelopmental CNVs appear to be associated with depression, extending the spectrum of clinical phenotypes that are associated with CNV carrier status.
Alzheimer's disease (AD) is a devastating neurodegenerative condition with significant genetic heritability. Several genes have been implicated in the onset of AD with the apolipoprotein E (APOE) ...gene being the strongest single genetic risk loci. Evidence suggests that the effect of APOE alters with age during disease progression. Here, we aim to investigate the impact of APOE and other variants outside the APOE region on AD risk in younger and older participants. Using data from both the Alzheimer's Disease Neuroimaging Initiative and the UK Biobank, we computed the polygenic risk score of each individual informed by the latest genetic study from the International Genomics of Alzheimer's Project. Our analysis showed that the effect of APOE on the disease risk is greater in younger participants and reduces as participants' age increases. Our findings indicate the increased impact of polygenic risk score as participants' age increases. Therefore, AD in older individuals can potentially be triggered by the cumulative effect of genes which are outside the APOE region.
•Polygenic risk score analysis was used in ADNI and the UK Biobank data sets.•APOE's effect on Alzheimer's disease risk was greater in the younger group (age<80 years).•Genes outside APOE region could trigger Alzheimer's disease in older ages (age≥80 years).•No considerable reduction of APOE ε4 alleles in ages less than and more than 80 years old.•Age-specific genetic scores could aid in clinical trials and personalized medicine.