文献速递头颈与中枢神经肿瘤

DeepClassPathway:使用可解释深度学习进行分子通路感知分类

DeepClassPathway: Molecular pathway aware classification using explainable deep learning.
PubMed 文献 · Eur J Cancer · 2022-09-30

研究开发了可解释卷积神经网络分类器DeepClassPathway,基于264例(13% HPV阳性)头颈鳞癌转录组数据训练、85例(25% HPV阳性)测试,将分子通路转化为二维树图以预测HPV状态并识别驱动决策的通路;CNN集成在全部树图变体上达到ROC-AUC 0.96、PR-AUC 0.90,并量化出对HPV阳性与阴性患者最具信息量的通路。

要点

  • CNN集成在全部树图变体上分类HPV状态的ROC-AUC为0.96、PR-AUC为0.90。
  • 对HPV阳性分类最具信息量的通路为KRAS、精子发生、胆汁酸代谢和炎症信号通路。
  • 对HPV阴性分类最具信息量的通路为MYC靶点、上皮-间质转化和蛋白分泌通路。

核心结论

  • 研究开发并应用了一种可解释CNN分类方法,对肿瘤队列转录组数据实现分类,同时考虑分子通路在个体层面决策中的重要性。

研究设计与样本

  • 研究提出可解释卷积神经网络分类器DeepClassPathway,在264例(13% HPV阳性)头颈鳞癌患者转录组数据上训练,在85例(25% HPV阳性)患者上测试。
  • 转录组数据转化为代表分子通路的二维树图,使用Grad-CAM显著性量化各通路对个体CNN决策的贡献,并通过在二维图像内打乱通路评估模型稳定性。

主要发现

  • CNN集成的分类性能在所有树图变体上达到ROC-AUC 0.96、PR-AUC 0.90。
  • 平均通路显著性热图量化一致识别KRAS、精子发生、胆汁酸代谢和炎症信号通路为分类HPV阳性患者最具信息量的四条通路。
  • 平均通路显著性热图量化识别MYC靶点、上皮-间质转化和蛋白分泌通路为分类HPV阴性患者最具信息量的通路。

局限性

  • 原文未报告局限性。

以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。

原文

OBJECTIVE: HPV-associated head and neck cancer is correlated with favorable prognosis; however, its underlying biology is not fully understood. We propose an explainable convolutional neural network (CNN) classifier, DeepClassPathway, that predicts HPV-status and allows patient-specific identification of molecular pathways driving classifier decisions. METHODS: The CNN was trained to classify HPV-status on transcriptome data from 264 (13% HPV-positive) and tested on 85 (25% HPV-positive) head and neck squamous carcinoma patients after transformation into 2D-treemaps representing molecular pathways. Grad-CAM saliency was used to quantify pathways contribution to individual CNN decisions. Model stability was assessed by shuffling pathways within 2D-images. RESULTS: The classification performance of the CNN-ensembles achieved ROC-AUC/PR-AUC of 0.96/0.90 for all treemap variants. Quantification of the averaged pathway saliency heatmaps consistently identified KRAS, spermatogenesis, bile acid metabolism, and inflammation signaling pathways as the four most informative for classifying HPV-positive patients and MYC targets, epithelial-mesenchymal transition, and protein secretion pathways for HPV-negative patients. CONCLUSION: We have developed and applied an explainable CNN classification approach to transcriptome data from an oncology cohort with typical sample size that allows classification while accounting for the importance of molecular pathways in individual-level decisions.

来源:Eur J Cancer,查看原文。版权归原作者所有。

引用

  1. Lombardo E, Hess J, Kurz C, et al. DeepClassPathway: Molecular pathway aware classification using explainable deep learning. Eur J Cancer. 2022. PMID: 36191385. doi:10.1016/j.ejca.2022.08.033

供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。