There are many active research projects accessing and applying shared ADNI data. Use the search above to find specific research focuses on the active ADNI investigations. This information is requested annually as a requirement for data access.
| Principal Investigator | |
| Principal Investigator's Name: | Marco Fernandes |
| Institution: | University of Oxford |
| Department: | Department of Psychiatry |
| Country: | |
| Proposed Analysis: | Aim: Investigating the association of gut microbiome and metabolome with psychiatric disorders using aggregated datasets. 1. We will employ traditional machine learning methods, including an array of clustering algorithms (centroid-based, connectivity-based, density-based and distribution-based), classification algorithms (random forests and support-vector machines), and state-of-art artificial neural networks to stratify disease subtypes and to predict response to treatment using metabolomic signatures. 2. Integration of different layers of information (molecular data, phenotypic and imaging features) from multi-Omics sources using graph-based approaches, such as network analysis using prior connectivity information from the Human interactome together with de novo data. This will potentially unravel by evaluating the network structure through applying random walks on a graph with Markov Chains algorithm, a set of ranked molecular pathways, network hubs (e.g. genes, metabolites, imaging feature) and network modules which discriminate (dependent upon the comparison used) disease subtypes or treatment modalities. |
| Additional Investigators | |
| Investigator's Name: | Liu Shi |
| Proposed Analysis: | Aim: Investigating the association of gut microbiome and metabolome with psychiatric disorders using aggregated datasets. 1. We will employ traditional machine learning methods, including an array of clustering algorithms (centroid-based, connectivity-based, density-based and distribution-based), classification algorithms (random forests and support-vector machines), and state-of-art artificial neural networks to stratify disease subtypes and to predict response to treatment using metabolomic signatures. 2. Integration of different layers of information (molecular data, phenotypic and imaging features) from multi-Omics sources using graph-based approaches, such as network analysis using prior connectivity information from the Human interactome together with de novo data. This will potentially unravel by evaluating the network structure by applying random walks on a graph with Markov Chains algorithm, a set of ranked molecular pathways, network hubs (e.g. genes, metabolites, imaging feature) and network modules which discriminate (dependent upon the comparison used) disease subtypes or treatment modalities. |
| Investigator's Name: | Alejo Nevado-Holgado |
| Proposed Analysis: | Aim: Investigating the association of gut microbiome and metabolome with psychiatric disorders using aggregated datasets. 1. We will employ traditional machine learning methods, including an array of clustering algorithms (centroid-based, connectivity-based, density-based and distribution-based), classification algorithms (random forests and support-vector machines), and state-of-art artificial neural networks to stratify disease subtypes and to predict response to treatment using metabolomic signatures. 2. Integration of different layers of information (molecular data, phenotypic and imaging features) from multi-Omics sources using graph-based approaches, such as network analysis using prior connectivity information from the Human interactome together with de novo data. This will potentially unravel by evaluating the network structure by applying random walks on a graph with Markov Chains algorithm, a set of ranked molecular pathways, network hubs (e.g. genes, metabolites, imaging feature) and network modules which discriminate (dependent upon the comparison used) disease subtypes or treatment modalities. |
| Investigator's Name: | William Sproviero |
| Proposed Analysis: | GWAS + WGS quality control (QC) and benchmark of the generated models. |
| Investigator's Name: | Daisy Sproviero |
| Proposed Analysis: | Generate linear mixed models (LMX) using the NMR-Nightingale longitudinal data, giving emphasis to gender-specific differences among a collection of disease phenotypes. |
