Clinical Validation of Whole Genome Methylation Profiling Classifier for Central Nervous System Tumors

Explore a comprehensive 35-minute conference talk from the Cancer Genomics Consortium's 2022 Annual Meeting, focusing on the clinical validation of whole genome methylation profiling classifiers for central nervous system tumors. Delve into the intricacies of sample workflow, methylation data analysis, and classification techniques. Learn about the normalization process, beta value interpretation, and the importance of training data in feature selection. Discover how dimensionality reduction impacts results and examine concordant data validation methods. Investigate typical cases through Disney Plots and Copy Number Plots, and engage in a thought-provoking discussion on the clinical applications of this technology. Gain insights into potential uses for constitutional disorders and explore the recalibration of probability scores. Conclude with a Q&A session addressing the role of pathologists in sample classification and the sharing of classifiers within the clinical laboratory setting.

Introduction
Disclosures
Sample Workflow
methylation data
normalization
beta value
classification
training data
feature selection
training classifier
dimensionality reduction
results
concordant data
validation
correlation
interference
Typical case
Disney Plot
Copy Number Plot
Discussion
Conclusions
Acknowledgements
Questions
Clinical Laboratory
Sharing the Classifier
Clinical Application
Constitutional Disorders
Additional Questions
Probability Score Recalibration
Can Pathologists Determine Classification of a Sample