Methylation data for "Rapid brain tumor classification from sparse epigenomic data"

Although the intraoperative, molecular differential diagnosis of the approximately one hundred different brain tumor entities described to date has been a goal of neuropathology in the last decade, this has not yet been achieved in a clinically relevant time frame of less than one hour after biopsy collection. Recent advances in third-generation sequencing technologies have brought this once-elusive goal within reach. However, established machine learning techniques rely on concepts and methods, impractical for live diagnostic workflows in clinical applications. Here, we present MethyLYZR, a Naïve Bayesian framework enabling fully tractable live classification of cancer epigenomes. MethyLYZR can be run in parallel with an ongoing Nanopore experiment with negligible computational cost and provides clinically relevant and accurate cancer classification results within 15 minutes of sequencing. Therefore, only the time required for DNA extraction and the Nanopore sequencer's maximum parallel throughput remain limiting factors for even faster time-to-results. We demonstrate the potential utility of the MethyLYZR framework not only for the neurosurgical intraoperative use case but also for other oncologic indications and cell-free DNA from liquid biopsies. This dataset provides methylation data from ONT and PacBio sequencing in feather file format.尽管近十年来，神经病理学领域始终将术中分子鉴别诊断目前已报道的近百种不同脑肿瘤实体作为核心目标，但截至目前，尚未能在活检取材后一小时这一临床相关时限内完成该诊断。第三代测序技术的新近突破，让这一曾遥不可及的目标逐渐触手可及。然而，现有成熟的机器学习技术所依托的理念与方法，并不适配临床场景中的实时诊断工作流。在此背景下，我们推出MethyLYZR——一种可实现癌症表观基因组完全可解释实时分类的朴素贝叶斯（Naïve Bayesian）框架。该工具可与正在运行的纳米孔（Nanopore）测序实验并行运行，计算成本可忽略不计，且能在测序完成后15分钟内输出符合临床要求的精准癌症分类结果。因此，即便要进一步缩短结果产出时长，仅剩DNA提取所需时长与纳米孔测序仪的最大并行通量两个限制因素。我们验证了MethyLYZR框架的应用潜力：其不仅可用于神经外科术中诊断场景，还可拓展至其他肿瘤学适应证以及来自液体活检的无细胞DNA（cell-free DNA）分析。 本数据集提供了采用ONT与PacBio测序技术生成的甲基化组数据，文件格式为Feather。