Software application for clustering for lipidomics and metabolomics

Unsupervised cluster analysis is one of the main steps in mining metabolomics and lipidomics data used to explain and group both samples and features. In this project we will work on building software applications in R or Python with RShiny GUI that makes several clustering algorithms easily accessible to experimentalists in metabolomics, lipidomics as well as other omics fields. Software will be posted on https://complimet.ca

Internship period: full time 2 month; part time 4 months. Approval from student graduate supervisor is required. All work will be done online.

This project is a collaboration between the National Research Council of Canada (NRC) and the University of Ottawa.

Spectral clustering for image analysis for MRI

In this project we will build algorithms and a Web application for spectral clustering combined with fuzzy clustering for image segmentation for further selection of significant regions or features (i.e. voxels). In this approach, MRI images will be perceived as a graph of connected voxels and spectral clustering will aim to choose graph subgroups defining regions of related behaviour. The advantage of spectral clustering in this context is that it does not require a user predefined number of clusters. In order to allow belonging of voxels (here seen as graph nodes) to multiple clusters we will combine spectral clustering with fuzzy clustering approach. We will build this approach for analysis of ADNI MRI data as a test set and aim towards development of a Web application posted on https://complimet.ca.

Internship period: full time 2 month; part time 4 months. Approval from student graduate supervisor is required. All work will be done online.

This project is a collaboration between the National Research Council of Canada (NRC) and the University of Ottawa.

Automated method for quantification of MRS data

As part of collaboration between NRC and Carleton University, we have developed methods for NMR quantum-theory based quantification methods for NMR metabolite measurements. The approach used in this internship will be adapted for use in MRS analysis for faster and more accurate quantification possibly developed for on-line application. Software is written in Julia but can be linked to Python, Matlab or R or RShiny for further development. We will build this approach for analysis of ADNI MRS data as a test set and aim towards development of a Web application posted on https://complimet.ca.

Internship period: full time 2 month; part time 4 months. Approval from student graduate supervisor is required. All work will be done online.

Project is a collaboration between the National Research Council of Canada (NRC), Carleton University, and the University of Ottawa.

Online application for previously developed AI method for ranking mouse Nest building performance

Mouse nesting test is one of the standard approaches for study of behavioral changes in dementia research. As a collaboration between University of Ottawa and NRC we have developed an AI algorithm for automated score assignment from mouse nest experiment from images. Further training and hyperparameter optimization is required as well as the development of RShiny GUI for online application that will allow for further utilization and method optimization. Student will work collaboratively with NRC team on the development of GUI and linking it to the existing Python code and development of Web application posted on https://complimet.ca.

Internship period: full time 2 month; part time 4 months. Approval from student graduate supervisor is required. All work will be done online.

This project is a collaboration between the National Research Council of Canada (NRC) and the University of Ottawa.