Multi-modality machine learning predicting Parkinson's disease

Diagnosis of Parkinson's disease based on voice signals using SHAP and hard voting ensemble method

Parkinson's disease (PD) is the second most common progressive neurological condition after Alzheimer's. The significant number of individuals afflicted with this illness makes it essential to develop a method to diagnose the conditions in their early phases. PD is typically identified from motor symptoms or via other Neuroimaging techniques. Expensive, time-consuming, and unavailable to the general public, these methods are not very accurate. Another issue to be addressed is the black-box nature of machine learning methods that needs interpretation. These issues encourage us to develop a novel technique using Shapley additive explanations (SHAP) and Hard Voting Ensemble Method based on voice signals to diagnose PD more accurately. Another purpose of this study is to interpret the output of the model and determine the most important features in diagnosing PD. The present article uses Pearson Correlation Coefficients to understand the relationship between input features and the output. Input features with high correlation are selected and then classified by the Extreme Gradient Boosting, Light Gradient Boosting Machine, Gradient Boosting, and Bagging. Moreover, the weights in Hard Voting Ensemble Method are determined based on the performance of the mentioned classifiers. At the final stage, it uses SHAP to determine the most important features in PD diagnosis. The effectiveness of the proposed method is validated using 'Parkinson Dataset with Replicated Acoustic Features' from the UCI machine learning repository. It has achieved an accuracy of 85.42%. The findings demonstrate that the proposed method outperformed state-of-the-art approaches and can assist physicians in diagnosing Parkinson's cases.

Deep Learning for Parkinson's Disease Diagnosis: A Graph Neural Network (GNN) Based Classification Approach with Graph Wavelet Transform (GWT) Using Protein-Peptide Datasets

Background: An important neurological disorder of Parkinson's Disease (PD) is characterized by motor and non-motor activity of the patients. Empirical condition of the patient: PD assessment uses the Movement Disorder Society Unified Parkinson's Rating Scale part III (MDS-UPDRS-III) measures for identifying the prediction of PD. Due to the unstable value of the measurement, the PD prediction and tracking lead to a lower prediction rate. Methods: To overcome this limitation, this paper proposed the Graph Wavelet Transform (GWT) based weighted feature extraction along with the Graph Neutral Network (GNN) classification. The main contribution of this research is (i) The weighted correlation between the data is calculated by GWT for effective prediction of PD. (ii) Machine learning algorithms were trained to predict Parkinson's disease based on these patterns. In this research, we developed a new model called Graph Neural Network (GNN) to predict PD tremors' MDS-UPDRS-III score using input data. To strengthen PD research and enable the construction of individualized treatment plans, these linked networks work together to methodically examine the data and find significant discoveries. Results: The proposed approach for predicting PD severity (motor- and MDS_UPDRS) has a mean squared error of 0.1796 and a root mean squared error of 0.2845, according to the experimental data. The prediction accuracy is increased by 27.66%, 54.11%, and 0.71%, correspondingly, when compared with the most effective State-of-the-Art methods of DNN, ANFIS + SVR, and Mixed MLP models. Conclusion: In conclusion, this proves that the proposed strategy is more effective at making predictions.

Non-Negative Matrix Tri-Factorization for Representation Learning in Multi-Omics Datasets with Applications to Drug Repurposing and Selection

The vast corpus of heterogeneous biomedical data stored in databases, ontologies, and terminologies presents a unique opportunity for drug design. Integrating and fusing these sources is essential to develop data representations that can be analyzed using artificial intelligence methods to generate novel drug candidates or hypotheses. Here, we propose Non-Negative Matrix Tri-Factorization as an invaluable tool for integrating and fusing data, as well as for representation learning. Additionally, we demonstrate how representations learned by Non-Negative Matrix Tri-Factorization can effectively be utilized by traditional artificial intelligence methods. While this approach is domain-agnostic and applicable to any field with vast amounts of structured and semi-structured data, we apply it specifically to computational pharmacology and drug repurposing. This field is poised to benefit significantly from artificial intelligence, particularly in personalized medicine. We conducted extensive experiments to evaluate the performance of the proposed method, yielding exciting results, particularly compared to traditional methods. Novel drug-target predictions have also been validated in the literature, further confirming their validity. Additionally, we tested our method to predict drug synergism, where constructing a classical matrix dataset is challenging. The method demonstrated great flexibility, suggesting its applicability to a wide range of tasks in drug design and discovery.

Interpretable video-based tracking and quantification of parkinsonism clinical motor states

Quantification of motor symptom progression in Parkinson's disease (PD) patients is crucial for assessing disease progression and for optimizing therapeutic interventions, such as dopaminergic medications and deep brain stimulation. Cumulative and heuristic clinical experience has identified various clinical signs associated with PD severity, but these are neither objectively quantifiable nor robustly validated. Video-based objective symptom quantification enabled by machine learning (ML) introduces a potential solution. However, video-based diagnostic tools often have implementation challenges due to expensive and inaccessible technology, and typical "black-box" ML implementations are not tailored to be clinically interpretable. Here, we address these needs by releasing a comprehensive kinematic dataset and developing an interpretable video-based framework that predicts high versus low PD motor symptom severity according to MDS-UPDRS Part III metrics. This data driven approach validated and robustly quantified canonical movement features and identified new clinical insights, not previously appreciated as related to clinical severity, including pinkie finger movements and lower limb and axial features of gait. Our framework is enabled by retrospective, single-view, seconds-long videos recorded on consumer-grade devices such as smartphones, tablets, and digital cameras, thereby eliminating the requirement for specialized equipment. Following interpretable ML principles, our framework enforces robustness and interpretability by integrating (1) automatic, data-driven kinematic metric evaluation guided by pre-defined digital features of movement, (2) combination of bi-domain (body and hand) kinematic features, and (3) sparsity-inducing and stability-driven ML analysis with simple-to-interpret models. These elements ensure that the proposed framework quantifies clinically meaningful motor features useful for both ML predictions and clinical analysis.

Transcriptional pathobiology and multi-omics predictors for Parkinson's disease

Early diagnosis and biomarker discovery to bolster the therapeutic pipeline for Parkinson's disease (PD) are urgently needed. In this study, we leverage the large-scale whole-blood total RNA-seq dataset from the Accelerating Medicine Partnership in Parkinson's Disease (AMP PD) program to identify PD-associated RNAs, including both known genes and novel circular RNAs (circRNA) and enhancer RNAs (eRNAs). There were 1,111 significant marker RNAs, including 491 genes, 599 eRNAs, and 21 circRNAs, that were first discovered in the PPMI cohort (FDR < 0.05) and confirmed in the PDBP/BioFIND cohorts (nominal p < 0.05). Functional enrichment analysis showed that the PD-associated genes are involved in neutrophil activation and degranulation, as well as the TNF-alpha signaling pathway. We further compare the PD-associated genes in blood with those in post-mortem brain dopamine neurons in our BRAINcode cohort. 44 genes show significant changes with the same direction in both PD brain neurons and PD blood, including neuroinflammation-associated genes IKBIP, CXCR2, and NFKBIB. Finally, we built a novel multi-omics machine learning model to predict PD diagnosis with high performance (AUC = 0.89), which was superior to previous studies and might aid the decision-making for PD diagnosis in clinical practice. In summary, this study delineates a wide spectrum of the known and novel RNAs linked to PD and are detectable in circulating blood cells in a harmonized, large-scale dataset. It provides a generally useful computational framework for further biomarker development and early disease prediction.

A Systematic Review of Genetics- and Molecular-Pathway-Based Machine Learning Models for Neurological Disorder Diagnosis

The process of identification and management of neurological disorder conditions faces challenges, prompting the investigation of novel methods in order to improve diagnostic accuracy. In this study, we conducted a systematic literature review to identify the significance of genetics- and molecular-pathway-based machine learning (ML) models in treating neurological disorder conditions. According to the study's objectives, search strategies were developed to extract the research studies using digital libraries. We followed rigorous study selection criteria. A total of 24 studies met the inclusion criteria and were included in the review. We classified the studies based on neurological disorders. The included studies highlighted multiple methodologies and exceptional results in treating neurological disorders. The study findings underscore the potential of the existing models, presenting personalized interventions based on the individual's conditions. The findings offer better-performing approaches that handle genetics and molecular data to generate effective outcomes. Moreover, we discuss the future research directions and challenges, emphasizing the demand for generalizing existing models in real-world clinical settings. This study contributes to advancing knowledge in the field of diagnosis and management of neurological disorders.

A Machine Learning Approach for Early Identification of Prodromal Parkinson's Disease

Parkinson's disease (PD) affects approximately 6 million people worldwide. Data analysis of early PD symptoms using machine learning (ML) models may provide an inexpensive, non-invasive, and simple method for the remote diagnosis of early PD. The aim of this project was to analyze voice, computer keystrokes, spiral drawings, and gait data involving PD patients and controls available in public databases using ML models and identify early PD characteristics that are more pronounced than others. An ML model was developed using Random Forest to analyze existing clinical data for PD patients, prodromal PD patients with REM (rapid eye movement) sleep behavior disorder (RBD) symptoms, and non-PD healthy controls. We reviewed and collected data from the UCI (University of California Irvine) Machine Learning Repository, PPMI (Parkinson's Progression Markers Initiative), and Kaggle databases. ML analysis was carried out on voice samples in PD and RBD patients, computer keystroke data, spiral drawings, and gait datasets. The ML prediction model developed may be helpful in improving risk prediction for PD, enabling early intervention and resource prioritization. The ML study suggests that voice analysis is the most robust test, followed by computer keystroke data, spiral drawings, and gait analysis, in that order. Voice is affected even in RBD patients, revealing that it is a sensitive and early measure of prodromal PD. The low accuracy of the analysis indicates that several PD-positive samples may remain undetected and unclassified. Combining all four features, that is, voice analysis, computer keystroke data, spiral drawings, and gait analysis, may improve the overall accuracy.

Testing the generalizability and effectiveness of deep learning models among clinics: sperm detection as a pilot study

BACKGROUND

Deep learning has been increasingly investigated for assisting clinical in vitro fertilization (IVF). The first technical step in many tasks is to visually detect and locate sperm, oocytes, and embryos in images. For clinical deployment of such deep learning models, different clinics use different image acquisition hardware and different sample preprocessing protocols, raising the concern over whether the reported accuracy of a deep learning model by one clinic could be reproduced in another clinic. Here we aim to investigate the effect of each imaging factor on the generalizability of object detection models, using sperm analysis as a pilot example.

METHODS

Ablation studies were performed using state-of-the-art models for detecting human sperm to quantitatively assess how model precision (false-positive detection) and recall (missed detection) were affected by imaging magnification, imaging mode, and sample preprocessing protocols. The results led to the hypothesis that the richness of image acquisition conditions in a training dataset deterministically affects model generalizability. The hypothesis was tested by first enriching the training dataset with a wide range of imaging conditions, then validated through internal blind tests on new samples and external multi-center clinical validations.

RESULTS

Ablation experiments revealed that removing subsets of data from the training dataset significantly reduced model precision. Removing raw sample images from the training dataset caused the largest drop in model precision, whereas removing 20x images caused the largest drop in model recall. by incorporating different imaging and sample preprocessing conditions into a rich training dataset, the model achieved an intraclass correlation coefficient (ICC) of 0.97 (95% CI: 0.94-0.99) for precision, and an ICC of 0.97 (95% CI: 0.93-0.99) for recall. Multi-center clinical validation showed no significant differences in model precision or recall across different clinics and applications.

CONCLUSIONS

The results validated the hypothesis that the richness of data in the training dataset is a key factor impacting model generalizability. These findings highlight the importance of diversity in a training dataset for model evaluation and suggest that future deep learning models in andrology and reproductive medicine should incorporate comprehensive feature sets for enhanced generalizability across clinics.

Insights into Ancestral Diversity in Parkinsons Disease Risk: A Comparative Assessment of Polygenic Risk Scores

Objectives To evaluate and compare different polygenic risk score (PRS) models in predicting Parkinsons disease (PD) across diverse ancestries, focusing on identifying the most suitable approach for each population and potentially contributing to equitable advancements in precision medicine. Methods We constructed a total of 105 PRS across individual level data from seven diverse ancestries. First, a cross-ancestry conventional PRS comparison was implemented by utilizing the 90 known European risk loci with weighted effects from four independent summary statistics including European, East Asian, Latino/Admixed American, and African/Admixed. These models were adjusted by sex, age, and principal components (28 PRS) and by sex, age, and percentage of admixture (28 PRS) for comparison. Secondly, a novel and refined multi-ancestry best-fit PRS approach was then applied across the seven ancestries by leveraging multi-ancestry meta-analyzed summary statistics and using a p-value thresholding approach (49 PRS) to enhance prediction applicability in a global setting. Results European-based PRS models predicted disease status across all ancestries to differing degrees of accuracy. Ashkenazi Jewish had the highest Odds Ratio (OR): 1.96 (95% CI: 1.69-2.25, p < 0.0001) with an AUC (Area Under the Curve) of 68%. Conversely, the East Asian population, despite having fewer predictive variants (84 out of 90), had an OR of 1.37 (95% CI: 1.32-1.42) and an AUC of 62%, illustrating the cross-ancestry transferability of this model. Lower OR alongside broader confidence intervals were observed in other populations, including Africans (OR =1.38, 95% CI: 1.12-1.63, p=0.001). Adjustment by percentage of admixture did not outperform principal components. Multi-ancestry best-fit PRS models improved risk prediction in European, Ashkenazi Jewish, and African ancestries, yet didn't surpass conventional PRS in admixed populations such as Latino/American admixed and African admixed populations. Interpretation The present study represents a novel and comprehensive assessment of PRS performance across seven ancestries in PD, highlighting the inadequacy of a 'one size fits all' approach in genetic risk prediction. We demonstrated that European based PD PRS models are partially transferable to other ancestries and could be improved by a novel best-fit multi-ancestry PRS, especially in non-admixed populations.

Towards a Global View of Parkinson's Disease Genetics

Parkinson's disease (PD) is a global health challenge, yet historically studies of PD have taken place predominantly in European populations. Recent genetics research conducted in non-European populations has revealed novel population-specific genetic loci linked to PD risk, highlighting the importance of studying PD globally. These insights have broadened our understanding of PD etiology, which is crucial for developing disease-modifying interventions. This review comprehensively explores the global genetic landscape of PD, emphasizing the scientific rationale for studying underrepresented populations. It underscores challenges, such as genotype-phenotype heterogeneity and inclusion difficulties for non-European participants, emphasizing the ongoing need for diverse and inclusive research in PD. ANN NEUROL 2024;95:831-842.

Prioritization of genes for translation: a computational approach

INTRODUCTION

Gene identification for genetic diseases is critical for the development of new diagnostic approaches and personalized treatment options. Prioritization of gene translation is an important consideration in the molecular biology field, allowing researchers to focus on the most promising candidates for further investigation.

AREAS COVERED

In this paper, we discussed different approaches to prioritize genes for translation, including the use of computational tools and machine learning algorithms, as well as experimental techniques such as knockdown and overexpression studies. We also explored the potential biases and limitations of these approaches and proposed strategies to improve the accuracy and reliability of gene prioritization methods. Although numerous computational methods have been developed for this purpose, there is a need for computational methods that incorporate tissue-specific information to enable more accurate prioritization of candidate genes. Such methods should provide tissue-specific predictions, insights into underlying disease mechanisms, and more accurate prioritization of genes.

EXPERT OPINION

Using advanced computational tools and machine learning algorithms to prioritize genes, we can identify potential targets for therapeutic intervention of complex diseases. This represents an up-and-coming method for drug development and personalized medicine.

Patient-specific game-based transfer method for Parkinson's disease severity prediction

Dysphonia is one of the early symptoms of Parkinson's disease (PD). Most existing methods use feature selection methods to find the optimal subset of voice features for all PD patients. Few have considered the heterogeneity between patients, which implies the need to provide specific prediction models for different patients. However, building the specific model faces the challenge of small sample size, which makes it lack generalization ability. Instance transfer is an effective way to solve this problem. Therefore, this paper proposes a patient-specific game-based transfer (PSGT) method for PD severity prediction. First, a selection mechanism is used to select PD patients with similar disease trends to the target patient from the source domain, which reduces the risk of negative transfer. Then, the contribution of the transferred subjects and their instances to the disease estimation of the target subject is fairly evaluated by the Shapley value, which improves the interpretability of the method. Next, the proportion of valid instances in the transferred subjects is determined, and the instances with higher contribution are transferred to further reduce the difference between the transferred instance subset and the target subject. Finally, the selected subset of instances is added to the training set of the target subject, and the extended data is fed into the random forest to improve the performance of the method. Parkinson's telemonitoring dataset is used to evaluate the feasibility and effectiveness. The mean values of mean absolute error, root mean square error, and volatility obtained by predicting motor-UPDRS and total-UPDRS for target patients are 1.59, 1.95, 1.56 and 1.98, 2.54, 1.94, respectively. Experiment results show that the PSGT has better performance in both prediction error and stability over compared methods.

Genetics in Parkinson's disease, state-of-the-art and future perspectives

BACKGROUND

Parkinson's disease (PD) is the second most common neurodegenerative disorder and is clinically characterized by the presence of motor (bradykinesia, rigidity, rest tremor and postural instability) and non-motor symptoms (cognitive impairment, autonomic dysfunction, sleep disorders, depression and hyposmia). The aetiology of PD is unknown except for a small but significant contribution of monogenic forms.

SOURCES OF DATA

No new data were generated or analyzed in support of this review.

AREAS OF AGREEMENT

Up to 15% of PD patients carry pathogenic variants in PD-associated genes. Some of these genes are associated with mendelian inheritance, while others act as risk factors. Genetic background influences age of onset, disease course, prognosis and therapeutic response.

AREAS OF CONTROVERSY

Genetic testing is not routinely offered in the clinical setting, but it may have relevant implications, especially in terms of prognosis, response to therapies and inclusion in clinical trials. Widely adopted clinical guidelines on genetic testing are still lacking and open to debate. Some new genetic associations are still awaiting confirmation, and selecting the appropriate genes to be included in diagnostic panels represents a difficult task. Finally, it is still under study whether (and to which degree) specific genetic forms may influence the outcome of PD therapies.

GROWING POINTS

Polygenic Risk Scores (PRS) may represent a useful tool to genetically stratify the population in terms of disease risk, prognosis and therapeutic outcomes.

AREAS TIMELY FOR DEVELOPING RESEARCH

The application of PRS and integrated multi-omics in PD promises to improve the personalized care of patients.

Exploring the application and challenges of fNIRS technology in early detection of Parkinson's disease

Background

Parkinson's disease (PD) is a prevalent neurodegenerative disorder that significantly benefits from early diagnosis for effective disease management and intervention. Despite advancements in medical technology, there remains a critical gap in the early and non-invasive detection of PD. Current diagnostic methods are often invasive, expensive, or late in identifying the disease, leading to missed opportunities for early intervention.

Objective

The goal of this study is to explore the efficiency and accuracy of combining fNIRS technology with machine learning algorithms in diagnosing early-stage PD patients and to evaluate the feasibility of this approach in clinical practice.

Methods

Using an ETG-4000 type near-infrared brain function imaging instrument, data was collected from 120 PD patients and 60 healthy controls. This cross-sectional study employed a multi-channel mode to monitor cerebral blood oxygen changes. The collected data were processed using a general linear model and β values were extracted. Subsequently, four types of machine learning models were developed for analysis: Support vector machine (SVM), K-nearest neighbors (K-NN), random forest (RF), and logistic regression (LR). Additionally, SHapley Additive exPlanations (SHAP) technology was applied to enhance model interpretability.

Results

The SVM model demonstrated higher accuracy in differentiating between PD patients and control group (accuracy of 85%, f1 score of 0.85, and an area under the ROC curve of 0.95). SHAP analysis identified the four most contributory channels (CH) as CH01, CH04, CH05, and CH08.

Conclusion

The model based on the SVM algorithm exhibited good diagnostic performance in the early detection of PD patients. Future early diagnosis of PD should focus on the Frontopolar Cortex (FPC) region.

Federated learning for multi-omics: A performance evaluation in Parkinson's disease

While machine learning (ML) research has recently grown more in popularity, its application in the omics domain is constrained by access to sufficiently large, high-quality datasets needed to train ML models. Federated learning (FL) represents an opportunity to enable collaborative curation of such datasets among participating institutions. We compare the simulated performance of several models trained using FL against classically trained ML models on the task of multi-omics Parkinson's disease prediction. We find that FL model performance tracks centrally trained ML models, where the most performant FL model achieves an AUC-PR of 0.876 ± 0.009, 0.014 ± 0.003 less than its centrally trained variation. We also determine that the dispersion of samples within a federation plays a meaningful role in model performance. Our study implements several open-source FL frameworks and aims to highlight some of the challenges and opportunities when applying these collaborative methods in multi-omics studies.

Federated Learning for multi-omics: a performance evaluation in Parkinson's disease

While machine learning (ML) research has recently grown more in popularity, its application in the omics domain is constrained by access to sufficiently large, high-quality datasets needed to train ML models. Federated Learning (FL) represents an opportunity to enable collaborative curation of such datasets among participating institutions. We compare the simulated performance of several models trained using FL against classically trained ML models on the task of multi-omics Parkinson's Disease prediction. We find that FL model performance tracks centrally trained ML models, where the most performant FL model achieves an AUC-PR of 0.876 ± 0.009, 0.014 ± 0.003 less than its centrally trained variation. We also determine that the dispersion of samples within a federation plays a meaningful role in model performance. Our study implements several open source FL frameworks and aims to highlight some of the challenges and opportunities when applying these collaborative methods in multi-omics studies.

Early-stage idiopathic Parkinson's disease is associated with reduced circular RNA expression

Neurodegeneration in Parkinson's disease (PD) precedes diagnosis by years. Early neurodegeneration may be reflected in RNA levels and measurable as a biomarker. Here, we present the largest quantification of whole blood linear and circular RNAs (circRNA) in early-stage idiopathic PD, using RNA sequencing data from two cohorts (PPMI = 259 PD, 161 Controls; ICICLE-PD = 48 PD, 48 Controls). We identified a replicable increase in TMEM252 and LMNB1 gene expression in PD. We identified novel differences in the expression of circRNAs from ESYT2, BMS1P1 and CCDC9, and replicated trends of previously reported circRNAs. Overall, using circRNA as a diagnostic biomarker in PD did not show any clear improvement over linear RNA, minimising its potential clinical utility. More interestingly, we observed a general reduction in circRNA expression in both PD cohorts, accompanied by an increase in RNASEL expression. This imbalance implicates the activation of an innate antiviral immune response and suggests a previously unknown aspect of circRNA regulation in PD.

Explainable machine learning for predicting conversion to neurological disease: Results from 52,939 medical records

Objective

This study assesses the application of interpretable machine learning modeling using electronic medical record data for the prediction of conversion to neurological disease.

Methods

A retrospective dataset of Cleveland Clinic patients diagnosed with Alzheimer's disease, amyotrophic lateral sclerosis, multiple sclerosis, or Parkinson's disease, and matched controls based on age, sex, race, and ethnicity was compiled. Individualized risk prediction models were created using eXtreme Gradient Boosting for each neurological disease at four timepoints in patient history. The prediction models were assessed for transparency and fairness.

Results

At timepoints 0-months, 12-months, 24-months, and 60-months prior to diagnosis, Alzheimer's disease models achieved the area under the receiver operating characteristic curve on a holdout test dataset of 0.794, 0.742, 0.709, and 0.645; amyotrophic lateral sclerosis of 0.883, 0.710, 0.658, and 0.620; multiple sclerosis of 0.922, 0.877, 0.849, and 0.781; and Parkinson's disease of 0.809, 0.738, 0.700, and 0.651, respectively.

Conclusions

The results demonstrate that electronic medical records contain latent information that can be used for risk stratification for neurological disorders. In particular, patient-reported outcomes, sleep assessments, falls data, additional disease diagnoses, and longitudinal changes in patient health, such as weight change, are important predictors.

Cell-free RNA signatures predict Alzheimer's disease

There is a need for affordable, scalable, and specific blood-based biomarkers for Alzheimer's disease that can be applied to a population level. We have developed and validated disease-specific cell-free transcriptomic blood-based biomarkers composed by a scalable number of transcripts that capture AD pathobiology even in the presymptomatic stages of the disease. Accuracies are in the range of the current CSF and plasma biomarkers, and specificities are high against other neurodegenerative diseases.

Prediction of Parkinson's Disease Using Machine Learning Methods

The detection of Parkinson's disease (PD) in its early stages is of great importance for its treatment and management, but consensus is lacking on what information is necessary and what models should be used to best predict PD risk. In our study, we first grouped PD-associated factors based on their cost and accessibility, and then gradually incorporated them into risk predictions, which were built using eight commonly used machine learning models to allow for comprehensive assessment. Finally, the Shapley Additive Explanations (SHAP) method was used to investigate the contributions of each factor. We found that models built with demographic variables, hospital admission examinations, clinical assessment, and polygenic risk score achieved the best prediction performance, and the inclusion of invasive biomarkers could not further enhance its accuracy. Among the eight machine learning models considered, penalized logistic regression and XGBoost were the most accurate algorithms for assessing PD risk, with penalized logistic regression achieving an area under the curve of 0.94 and a Brier score of 0.08. Olfactory function and polygenic risk scores were the most important predictors for PD risk. Our research has offered a practical framework for PD risk assessment, where necessary information and efficient machine learning tools were highlighted.

Artificial intelligence for dementia genetics and omics

Genetics and omics studies of Alzheimer's disease and other dementia subtypes enhance our understanding of underlying mechanisms and pathways that can be targeted. We identified key remaining challenges: First, can we enhance genetic studies to address missing heritability? Can we identify reproducible omics signatures that differentiate between dementia subtypes? Can high-dimensional omics data identify improved biomarkers? How can genetics inform our understanding of causal status of dementia risk factors? And which biological processes are altered by dementia-related genetic variation? Artificial intelligence (AI) and machine learning approaches give us powerful new tools in helping us to tackle these challenges, and we review possible solutions and examples of best practice. However, their limitations also need to be considered, as well as the need for coordinated multidisciplinary research and diverse deeply phenotyped cohorts. Ultimately AI approaches improve our ability to interrogate genetics and omics data for precision dementia medicine. HIGHLIGHTS: We have identified five key challenges in dementia genetics and omics studies. AI can enable detection of undiscovered patterns in dementia genetics and omics data. Enhanced and more diverse genetics and omics datasets are still needed. Multidisciplinary collaborative efforts using AI can boost dementia research.

Volumetric changes and clinical trajectories in Parkinson's disease: a prospective multicentric study

BACKGROUND

Longitudinal measures of structural brain changes using MRI in relation to clinical features and progression patterns in PD have been assessed in previous studies, but few were conducted in well-defined and large cohorts, including prospective clinical assessments of both motor and non-motor symptoms.

OBJECTIVE

We aimed to identify brain volumetric changes characterizing PD patients, and determine whether regional brain volumetric characteristics at baseline can predict motor, psycho-behavioral and cognitive evolution at one year in a prospective cohort of PD patients.

METHODS

In this multicentric 1 year longitudinal study, PD patients and healthy controls from the MPI-R2* cohort were assessed for demographical, clinical and brain volumetric characteristics. Distinct subgroups of PD patients according to motor, cognitive and psycho-behavioral evolution were identified at the end of follow-up.

RESULTS

One hundred and fifty PD patients and 73 control subjects were included in our analysis. Over one year, there was no significant difference in volume variations between PD and control subjects, regardless of the brain region considered. However, we observed a reduction in posterior cingulate cortex volume at baseline in PD patients with motor deterioration at one year (p = 0.017). We also observed a bilateral reduction of the volume of the amygdala (p = 0.015 and p = 0.041) and hippocampus (p = 0.015 and p = 0.053) at baseline in patients with psycho-behavioral deterioration, regardless of age, dopaminergic treatment and center.

CONCLUSION

Brain volumetric characteristics at baseline may predict clinical trajectories at 1 year in PD as posterior cingulate cortex atrophy was associated with motor decline, while amygdala and hippocampus atrophy were associated with psycho-behavioral decline.

Genetic risk factor clustering within and across neurodegenerative diseases

Overlapping symptoms and co-pathologies are common in closely related neurodegenerative diseases (NDDs). Investigating genetic risk variants across these NDDs can give further insight into disease manifestations. In this study we have leveraged genome-wide single nucleotide polymorphisms and genome-wide association study summary statistics to cluster patients based on their genetic status across identified risk variants for five NDDs (Alzheimer's disease, Parkinson's disease, amyotrophic lateral sclerosis, Lewy body dementia and frontotemporal dementia). The multi-disease and disease-specific clustering results presented here provide evidence that NDDs have more overlapping genetic aetiology than previously expected and how neurodegeneration should be viewed as a spectrum of symptomology. These clustering analyses also show potential subsets of patients with these diseases that are significantly depleted for any known common genetic risk factors suggesting environmental or other factors at work. Establishing that NDDs with overlapping pathologies share genetic risk loci, future research into how these variants might have different effects on downstream protein expression, pathology and NDD manifestation in general is important for refining and treating NDDs.

A review of emergent intelligent systems for the detection of Parkinson's disease

Parkinson's disease (PD) is a neurodegenerative disorder affecting people worldwide. The PD symptoms are divided into motor and non-motor symptoms. Detection of PD is very crucial and essential. Such challenges can be overcome by applying artificial intelligence to diagnose PD. Many studies have also proposed the implementation of computer-aided diagnosis for the detection of PD. This systematic review comprehensively analyzed all appropriate algorithms for detecting and assessing PD based on the literature from 2012 to 2023 which are conducted as per PRISMA model. This review focused on motor symptoms, namely handwriting dynamics, voice impairments and gait, multimodal features, and brain observation using single photon emission computed tomography, magnetic resonance and electroencephalogram signals. The significant challenges are critically analyzed, and appropriate recommendations are provided. The critical discussion of this review article can be helpful in today's PD community in such a way that it allows clinicians to provide proper treatment and timely medication.

High-Throughput CSF Proteomics and Machine Learning to Identify Proteomic Signatures for Parkinson Disease Development and Progression

BACKGROUND AND OBJECTIVES

This study aimed to identify CSF proteomic signatures characteristic of Parkinson disease (PD) and evaluate their clinical utility.

METHODS

This observational study used data from the Parkinson's Progression Markers Initiative (PPMI), which enrolled patients with PD, healthy controls (HCs), and non-PD participants carrying GBA1, LRRK2, and/or SNCA pathogenic variants (genetic prodromals) at international sites. Study participants were chosen from PPMI enrollees based on the availability of aptamer-based CSF proteomic data, quantifying 4,071 proteins, and classified as patients with PD without GBA1, LRRK2, and/or SNCA pathogenic variants (nongenetic PD), HCs, patients with PD carrying the aforementioned pathogenic variants (genetic PD), or genetic prodromals. Differentially expressed protein (DEP) analysis and the least absolute shrinkage and selection operator (LASSO) were applied to the data from nongenetic PD and HCs. Signatures characteristics of nongenetic PD were quantified as a PD proteomic score (PD-ProS), validated internally and then externally using data of 1,556 CSF proteins from the LRRK2 Cohort Consortium (LCC). We further tested the PD-ProS in genetic PD and genetic prodromals and examined associations with clinical progression.

RESULTS

Data from 279 patients with nongenetic PD (mean ± SD, age 62.0 ± 9.6 years; male 67.7%) and 141 HCs (age 60.5 ± 11.9 years; male 64.5%) were used for PD-ProS derivation. From 23 DEPs, LASSO determined weights of 14 DEPs for the PD-ProS (area under the curve [AUC] 0.83, 95% CI 0.78-0.87), validated in an independent internal validation cohort of 71 patients with nongenetic PD and 35 HCs (AUC 0.81, 95% CI 0.73-0.90). In the LCC, only 5 of the 14 DEPs were also measured. Notably, these 5 DEPs still distinguished 34 patients with nongenetic PD from 31 HCs with the same weights (AUC 0.75, 95% CI 0.63-0.87). Furthermore, the PD-ProS distinguished 258 patients with genetic PD from 365 genetic prodromals. Finally, regardless of genetic status, the PD-ProS independently predicted both cognitive and motor decline in PD (dementia, adjusted hazard ratio in the highest quintile [aHR-Q5] 2.8 [95% CI 1.6-5.0]; Hoehn and Yahr stage IV, aHR-Q5 2.1 [95% CI 1.1-4.0]).

DISCUSSION

By integrating high-throughput proteomics with machine learning, we identified PD-associated CSF proteomic signatures crucial for PD development and progression.

TRIAL REGISTRATION INFORMATION

ClinicalTrials.gov (NCT01176565). A link to the trial registry page is clinicaltrials.gov/ct2/show/NCT01141023.

CLASSIFICATION OF EVIDENCE

This study provides Class II evidence that the CSF proteome contains clinically important information regarding the development and progression of Parkinson disease that can be deciphered by a combination of high-throughput proteomics and machine learning.

The differential diagnosis value of radiomics-based machine learning in Parkinson's disease: a systematic review and meta-analysis

Background

In recent years, radiomics has been increasingly utilized for the differential diagnosis of Parkinson's disease (PD). However, the application of radiomics in PD diagnosis still lacks sufficient evidence-based support. To address this gap, we carried out a systematic review and meta-analysis to evaluate the diagnostic value of radiomics-based machine learning (ML) for PD.

Methods

We systematically searched Embase, Cochrane, PubMed, and Web of Science databases as of November 14, 2022. The radiomics quality assessment scale (RQS) was used to evaluate the quality of the included studies. The outcome measures were the c-index, which reflects the overall accuracy of the model, as well as sensitivity and specificity. During this meta-analysis, we discussed the differential diagnostic value of radiomics-based ML for Parkinson's disease and various atypical parkinsonism syndromes (APS).

Results

Twenty-eight articles with a total of 6,057 participants were included. The mean RQS score for all included articles was 10.64, with a relative score of 29.56%. The pooled c-index, sensitivity, and specificity of radiomics for predicting PD were 0.862 (95% CI: 0.833-0.891), 0.91 (95% CI: 0.86-0.94), and 0.93 (95% CI: 0.87-0.96) in the training set, and 0.871 (95% CI: 0.853-0.890), 0.86 (95% CI: 0.81-0.89), and 0.87 (95% CI: 0.83-0.91) in the validation set, respectively. Additionally, the pooled c-index, sensitivity, and specificity of radiomics for differentiating PD from APS were 0.866 (95% CI: 0.843-0.889), 0.86 (95% CI: 0.84-0.88), and 0.80 (95% CI: 0.75-0.84) in the training set, and 0.879 (95% CI: 0.854-0.903), 0.87 (95% CI: 0.85-0.89), and 0.82 (95% CI: 0.77-0.86) in the validation set, respectively.

Conclusion

Radiomics-based ML can serve as a potential tool for PD diagnosis. Moreover, it has an excellent performance in distinguishing Parkinson's disease from APS. The support vector machine (SVM) model exhibits excellent robustness when the number of samples is relatively abundant. However, due to the diverse implementation process of radiomics, it is expected that more large-scale, multi-class image data can be included to develop radiomics intelligent tools with broader applicability, promoting the application and development of radiomics in the diagnosis and prediction of Parkinson's disease and related fields.

Systematic review registration

https://www.crd.york.ac.uk/PROSPERO/display_record.php?RecordID=383197, identifier ID: CRD42022383197.

Application of Aligned-UMAP to longitudinal biomedical studies

High-dimensional data analysis starts with projecting the data to low dimensions to visualize and understand the underlying data structure. Several methods have been developed for dimensionality reduction, but they are limited to cross-sectional datasets. The recently proposed Aligned-UMAP, an extension of the uniform manifold approximation and projection (UMAP) algorithm, can visualize high-dimensional longitudinal datasets. We demonstrated its utility for researchers to identify exciting patterns and trajectories within enormous datasets in biological sciences. We found that the algorithm parameters also play a crucial role and must be tuned carefully to utilize the algorithm's potential fully. We also discussed key points to remember and directions for future extensions of Aligned-UMAP. Further, we made our code open source to enhance the reproducibility and applicability of our work. We believe our benchmarking study becomes more important as more and more high-dimensional longitudinal data in biomedical research become available.

Nomogram for Early Prediction of Parkinson's Disease Based on microRNA Profiles and Clinical Variables

BACKGROUND

Few efficient and simple models for the early prediction of Parkinson's disease (PD) exists.

OBJECTIVE

To develop and validate a novel nomogram for early identification of PD by incorporating microRNA (miRNA) expression profiles and clinical indicators.

METHODS

Expression levels of blood-based miRNAs and clinical variables from 1,284 individuals were downloaded from the Parkinson's Progression Marker Initiative database on June 1, 2022. Initially, the generalized estimating equation was used to screen candidate biomarkers of PD progression in the discovery phase. Then, the elastic net model was utilized for variable selection and a logistics regression model was constructed to establish a nomogram. Additionally, the receiver operating characteristic (ROC) curves, decision curve analysis (DCA), and calibration curves were utilized to evaluate the performance of the nomogram.

RESULTS

An accurate and externally validated nomogram was constructed for predicting prodromal and early PD. The nomogram is easy to utilize in a clinical setting since it consists of age, gender, education level, and transcriptional score (calculated by 10 miRNA profiles). Compared with the independent clinical model or 10 miRNA panel separately, the nomogram was reliable and satisfactory because the area under the ROC curve achieved 0.72 (95% confidence interval, 0.68-0.77) and obtained a superior clinical net benefit in DCA based on external datasets. Moreover, calibration curves also revealed its excellent prediction power.

CONCLUSION

The constructed nomogram has potential for large-scale early screening of PD based upon its utility and precision.

Identification and prediction of Parkinson's disease subtypes and progression using machine learning in two cohorts

The clinical manifestations of Parkinson's disease (PD) are characterized by heterogeneity in age at onset, disease duration, rate of progression, and the constellation of motor versus non-motor features. There is an unmet need for the characterization of distinct disease subtypes as well as improved, individualized predictions of the disease course. We used unsupervised and supervised machine learning methods on comprehensive, longitudinal clinical data from the Parkinson's Disease Progression Marker Initiative (n = 294 cases) to identify patient subtypes and to predict disease progression. The resulting models were validated in an independent, clinically well-characterized cohort from the Parkinson's Disease Biomarker Program (n = 263 cases). Our analysis distinguished three distinct disease subtypes with highly predictable progression rates, corresponding to slow, moderate, and fast disease progression. We achieved highly accurate projections of disease progression 5 years after initial diagnosis with an average area under the curve (AUC) of 0.92 (95% CI: 0.95 ± 0.01) for the slower progressing group (PDvec1), 0.87 ± 0.03 for moderate progressors, and 0.95 ± 0.02 for the fast-progressing group (PDvec3). We identified serum neurofilament light as a significant indicator of fast disease progression among other key biomarkers of interest. We replicated these findings in an independent cohort, released the analytical code, and developed models in an open science manner. Our data-driven study provides insights to deconstruct PD heterogeneity. This approach could have immediate implications for clinical trials by improving the detection of significant clinical outcomes. We anticipate that machine learning models will improve patient counseling, clinical trial design, and ultimately individualized patient care.

Genetics in parkinson's disease: From better disease understanding to machine learning based precision medicine

Parkinson's Disease (PD) is a neurodegenerative disorder with highly heterogeneous phenotypes. Accordingly, it has been challenging to robustly identify genetic factors associated with disease risk, prognosis and therapy response via genome-wide association studies (GWAS). In this review we first provide an overview of existing statistical methods to detect associations between genetic variants and the disease phenotypes in existing PD GWAS. Secondly, we discuss the potential of machine learning approaches to better quantify disease phenotypes and to move beyond disease understanding towards a better-personalized treatment of the disease.

Genomics transformer for diagnosing Parkinson's disease

Parkinson's disease (PD) is the second most common neurodegenerative disease and presents a complex etiology with genomic and environmental factors and no recognized cures. Genotype data, such as single nucleotide polymorphisms (SNPs), could be used as a prodromal factor for early detection of PD. However, the polygenic nature of PD presents a challenge as the complex relationships between SNPs towards disease development are difficult to model. Traditional assessment methods such as polygenic risk scores and machine learning approaches struggle to capture the complex interactions present in the genotype data, thus limiting their discriminative capabilities in diagnosis. On the other hand, deep learning models are better suited for this task. Nevertheless, they encounter difficulties of their own such as a lack of interpretability. To overcome these limitations, in this work, a novel transformer encoder-based model is introduced to classify PD patients from healthy controls based on their genotype. This method is designed to effectively model complex global feature interactions and enable increased interpretability through the learned attention scores. The proposed framework outperformed traditional machine learning models and multilayer perceptron (MLP) baseline models. Moreover, visualization of the learned SNP-SNP associations provides not only interpretability to the model but also valuable insights into the biochemical pathways underlying PD development, which are corroborated by pathway enrichment analysis. Our results suggest novel SNP interactions to be further studied in wet lab and clinical settings.

Polygenic Resilience Modulates the Penetrance of Parkinson Disease Genetic Risk Factors

OBJECTIVE

The aim of the current study is to understand why some individuals avoid developing Parkinson disease (PD) despite being at relatively high genetic risk, using the largest datasets of individual-level genetic data available.

METHODS

We calculated polygenic risk score to identify controls and matched PD cases with the highest burden of genetic risk for PD in the discovery cohort (International Parkinson's Disease Genomics Consortium, 7,204 PD cases and 9,412 controls) and validation cohorts (Comprehensive Unbiased Risk Factor Assessment for Genetics and Environment in Parkinson's Disease, 8,968 cases and 7,598 controls; UK Biobank, 2,639 PD cases and 14,301 controls; Accelerating Medicines Partnership-Parkinson's Disease Initiative, 2,248 cases and 2,817 controls). A genome-wide association study meta-analysis was performed on these individuals to understand genetic variation associated with resistance to disease. We further constructed a polygenic resilience score, and performed multimarker analysis of genomic annotation (MAGMA) gene-based analyses and functional enrichment analyses.

RESULTS

A higher polygenic resilience score was associated with a lower risk for PD (β = -0.054, standard error [SE] = 0.022, p = 0.013). Although no single locus reached genome-wide significance, MAGMA gene-based analyses nominated TBCA as a putative gene. Furthermore, we estimated the narrow-sense heritability associated with resilience to PD (h2  = 0.081, SE = 0.035, p = 0.0003). Subsequent functional enrichment analysis highlighted histone methylation as a potential pathway harboring resilience alleles that could mitigate the effects of PD risk loci.

INTERPRETATION

The present study represents a novel and comprehensive assessment of heritable genetic variation contributing to PD resistance. We show that a genetic resilience score can modify the penetrance of PD genetic risk factors and therefore protect individuals carrying a high-risk genetic burden from developing PD. ANN NEUROL 2022;92:270-278.

Establishing gene regulatory networks from Parkinson's disease risk loci

The latest meta-analysis of genome-wide association studies identified 90 independent variants across 78 genomic regions associated with Parkinson's disease, yet the mechanisms by which these variants influence the development of the disease remains largely elusive. To establish the functional gene regulatory networks associated with Parkinson's disease risk variants, we utilized an approach combining spatial (chromosomal conformation capture) and functional (expression quantitative trait loci) data. We identified 518 genes subject to regulation by 76 Parkinson's variants across 49 tissues, whicih encompass 36 peripheral and 13 CNS tissues. Notably, one-third of these genes were regulated via trans-acting mechanisms (distal; risk locus-gene separated by >1 Mb, or on different chromosomes). Of particular interest is the identification of a novel trans-expression quantitative trait loci-gene connection between rs10847864 and SYNJ1 in the adult brain cortex, highlighting a convergence between familial studies and Parkinson's disease genome-wide association studies loci for SYNJ1 (PARK20) for the first time. Furthermore, we identified 16 neurodevelopment-specific expression quantitative trait loci-gene regulatory connections within the foetal cortex, consistent with hypotheses suggesting a neurodevelopmental involvement in the pathogenesis of Parkinson's disease. Through utilizing Louvain clustering we extracted nine significant and highly intraconnected clusters within the entire gene regulatory network. The nine clusters are enriched for specific biological processes and pathways, some of which have not previously been associated with Parkinson's disease. Together, our results not only contribute to an overall understanding of the mechanisms and impact of specific combinations of Parkinson's disease variants, but also highlight the potential impact gene regulatory networks may have when elucidating aetiological subtypes of Parkinson's disease.

A Machine Learning Approach to Parkinson’s Disease Blood Transcriptomics

The increased incidence and the significant health burden associated with Parkinson’s disease (PD) have stimulated substantial research efforts towards the identification of effective treatments and diagnostic procedures. Despite technological advancements, a cure is still not available and PD is often diagnosed a long time after onset when irreversible damage has already occurred. Blood transcriptomics represents a potentially disruptive technology for the early diagnosis of PD. We used transcriptome data from the PPMI study, a large cohort study with early PD subjects and age matched controls (HC), to perform the classification of PD vs. HC in around 550 samples. Using a nested feature selection procedure based on Random Forests and XGBoost we reached an AUC of 72% and found 493 candidate genes. We further discussed the importance of the selected genes through a functional analysis based on GOs and KEGG pathways.