Childhood Cancer Data Initiative Annual Symposium (Abstract Registration): Submission #84

Abstract Submission for Poster Presentation
-------------------------------------------
Abstract Title:: Nanopore Sequencing for Pediatric CNS Tumor Classification
Abstract::
Central nervous system (CNS) tumors are the most common solid tumors diagnosed in children, making up about 17% of diagnosed childhood cancer cases. Most pediatric CNS tumor cases require a combination of imaging, histopathological, and molecular diagnoses to determine the tumor (sub)type. My project utilizes an emerging molecular diagnostic technology -- nanopore sequencing -- to quickly and accurately generate high-depth whole genome DNA methylation profiles for a diverse cohort of 150 pediatric CNS tumors, and I am currently using this data to train and validate a pediatric-specific deep learning classifier. As part of a clinical research collaboration, I tested the previously-developed machine learning classifier Sturgeon on high-depth nanopore DNA methylation data for 7 matched pairs of standard patient tissue and surgical tumor aspirate (n=14). For each sample, I demonstrated between 7-34x genomic coverage, classified the sample’s tumor (sub)type, and used in-house software to validate key subtype-defining features and identify tumor-specific copy number variation(s). Through this initial analysis, I was able to identify key shortcomings in existing classifier(s), including inappropriate labeling for pediatric samples, insufficient adjustment for low tumor purity samples, and confident misclassification of rare/unique tumors. After making adjustments to existing preprocessing methods, all 14 samples were then classified correctly with high confidence. Matched samples also showed strong pairwise similarities, reflective of interchangeable use of the materials. Together, I used nanopore DNA methylation profiling to demonstrate fast and accurate classification of CNS tumor aspirate and to highlight areas of improvement for the development of a pediatric-specific CNS tumor classifier. 

Authors::
1. First Name: Allison
   Middle Initial: A
   Last Name: Murray
   Organization: University of North Carolina Chapel Hill
2. First Name: Breanna
   Middle Initial: E
   Last Name: Mann
   Degree(s): Ph.D.
   Organization: University of North Carolina Chapel Hill
3. First Name: Andrew
   Middle Initial: B
   Last Name: Satterlee
   Degree(s): Ph.D.
   Organization: University of North Carolina Chapel Hill
4. First Name: David
   Middle Initial: E
   Last Name: Kram
   Degree(s): M.D., MCR
   Organization: University of North Carolina Chapel Hill
5. First Name: Jeremy
   Middle Initial: R
   Last Name: Wang
   Degree(s): Ph.D.
   Organization: University of North Carolina Chapel Hill

Presenting Author:: Allison A Murray
Institution:: University of North Carolina Chapel Hill
Email Address:: amurray1@unc.edu