Deep learning-based approach for addressing missing values in mass spec

• Masters by Coursework
• Honours

The objective of this project is to address the significant challenge of missing values encountered during data analysis in mass spectrometry. Missing values can occur randomly or non-randomly, potentially leading to biased results after imputation.

Our proposal aims to develop a robust tool for missing value handling, to facilitate more accurate, insightful, and reliable analysis in proteomics research domains . This will be achieved by investigating the feasibility of employing deep learning techniques (Khan,Information Sciences 2022 595:278) to uncover intricate patterns and relationships within complex data to handle missing values in mass spectrometry data. Extensive experimentation and rigorous validation will be required to comprehensively assess the performance and reliability of our proposed deep learning-based methods.

The Proteomics Facility at WEHI is dedicated to advancing proteomics and optimising data analysis. The Proteomics Facility is planning to utilise the power of deep learning techniques to improve our data analysis.

With a primary focus on optimising data analysis workflows, our research interests lie at the intersection of proteomics, bioinformatics, and statistics. Leveraging our expertise in these areas, we develop in-house pipeline for processing, interpreting, and visualising high-dimensional proteomics data.

Our lab operates within a collaborative structure, involving talented scientists working together to achieve precise and comprehensive analysis of complex datasets and ultimately contributing to advancements in the understanding of disease mechanisms and the development of targeted therapeutics.