Novel Alzheimer's disease subtypes identified using a data and knowledge driven strategy.

The population of adults with Alzheimer's disease (AD) varies in needs and outcomes. The heterogeneity of current AD diagnostic subgroups impedes the use of data analytics in clinical trial design and translation of findings into improved care. The purpose of this project was to define more clinically-homogeneous groups of AD patients and link clinical characteristics with biological markers. We used an innovative big data analysis strategy, the 3C strategy, that incorporates medical knowledge into the data analysis process. A large set of preprocessed AD Neuroimaging Initiative (ADNI) data was analyzed with 3C. The data analysis yielded 6 new disease subtypes, which differ from the assigned diagnosis types and present different patterns of clinical measures and potential biomarkers. Two of the subtypes, "Anosognosia dementia" and "Insightful dementia", differentiate between severe participants based on clinical characteristics and biomarkers. The "Uncompensated mild cognitive impairment (MCI)" subtype, demonstrates clinical, demographic and imaging differences from the "Affective MCI" subtype. Differences were also observed between the "Worried Well" and "Healthy" clusters. The use of data-driven analysis yielded sub-phenotypic clinical clusters that go beyond current diagnoses and are associated with biomarkers. Such homogenous sub-groups can potentially form the basis for enhancement of brain medicine research.

Hierarchical Data-Driven Analysis of Clinical Symptoms Among Patients With Parkinson's Disease.

Mutations in the LRRK2 and GBA genes are the most common inherited causes of Parkinson's disease (PD). Studies exploring phenotypic differences based on genetic status used hypothesis-driven data-gathering and statistical-analyses focusing on specific symptoms, which may influence the validity of the results. We aimed to explore phenotypic expression in idiopathic PD (iPD) patients, G2019S-LRRK2-PD, and GBA-PD using a data-driven approach, allowing screening of large numbers of features while controlling selection bias. Data was collected from 1525 Ashkenazi Jews diagnosed with PD from the Tel-Aviv Medical center; 161 G2019S-LRRK2-PD, 222 GBA-PD, and 1142 iPD (no G2019S-LRRK2 or any of the 7 AJ GBA mutations tested). Data included 771 measures: demographics, cognitive, physical and neurological functions, performance-based measures, and non-motor symptoms. The association of the genotypes with each of the measures was tested while accounting for age at motor symptoms onset, gender, and disease duration; p-values were reported and corrected in a hierarchical approach for an average over the selected measures false discovery rate control, resulting in 32 measures. GBA-PD presented with more severe symptoms expression while LRRK2-PD had more benign symptoms compared to iPD. GBA-PD presented greater cognitive and autonomic involvement, more frequent hyposmia and REM sleep behavior symptoms while these were less frequent among LRRK2-PD compared to iPD. Using a data-driven analytical approach strengthens earlier studies and extends them to portray a possible unique disease phenotype based on genotype among AJ PD. Such findings could help direct a more personalized therapeutic approach.

The Importance of Nonlinear Transformations Use in Medical Data Analysis.

BACKGROUND: The accumulation of data and its accessibility through easier-to-use platforms will allow data scientists and practitioners who are less sophisticated data analysts to get answers by using big data for many purposes in multiple ways. Data scientists working with medical data are aware of the importance of preprocessing, yet in many cases, the potential benefits of using nonlinear transformations is overlooked. OBJECTIVE: Our aim is to present a semi-automated approach of symmetry-aiming transformations tailored for medical data analysis and its advantages. METHODS: We describe 10 commonly encountered data types used in the medical field and the relevant transformations for each data type. Data from the Alzheimer's Disease Neuroimaging Initiative study, Parkinson's disease hospital cohort, and disease-simulating data were used to demonstrate the approach and its benefits. RESULTS: Symmetry-targeted monotone transformations were applied, and the advantages gained in variance, stability, linearity, and clustering are demonstrated. An open source application implementing the described methods was developed. Both linearity of relationships and increase of stability of variability improved after applying proper nonlinear transformation. Clustering simulated nonsymmetric data gave low agreement to the generating clusters (Rand value=0.681), while capturing the original structure after applying nonlinear transformation to symmetry (Rand value=0.986). CONCLUSIONS: This work presents the use of nonlinear transformations for medical data and the importance of their semi-automated choice. Using the described approach, the data analyst increases the ability to create simpler, more robust and translational models, thereby facilitating the interpretation and implementation of the analysis by medical practitioners. Applying nonlinear transformations as part of the preprocessing is essential to the quality and interpretability of results.

Clusters of male and female Alzheimer's disease patients in the Alzheimer's Disease Neuroimaging Initiative (ADNI) database.

This paper presents homogeneous clusters of patients, identified in the Alzheimer's Disease Neuroimaging Initiative (ADNI) data population of 317 females and 342 males, described by a total of 243 biological and clinical descriptors. Clustering was performed with a novel methodology, which supports identification of patient subpopulations that are homogeneous regarding both clinical and biological descriptors. Properties of the constructed clusters clearly demonstrate the differences between female and male Alzheimer's disease patient groups. The major difference is the existence of two male subpopulations with unexpected values of intracerebral and whole brain volumes.