Bioinformatics Seminar
Time: 11AM
Venue: Zoom Webinar
13 April 2021
Measuring similarity between cancer samples and experimental models
Sam LeeWEHI Bioinformatics
Analysis of high-throughput data from primary cancer samples or experimental models often involves an analysis of the molecular differences between two or more groups in a study. However, the analysis of differences between defined groups may not be enough to understand how individual samples relate to each other. In this confirmation talk, I will focus on how to evaluate the similarity of samples between and across datasets.
I'll present the framework I have developed to evaluate the performance of methods addressing this question and detail the results of benchmarking to assess the performance of different methods for aligning experimental models from separate datasets using bulk RNA-Seq data.
This initial evaluation framework and benchmarking study will provide the framework that I will go on to use to investigate how targeted feature engineering may improve our ability to match between experimental models and patient samples for which we have limited information.