Performance-weighted-voting model: an ensemble machine learning method for cancer type classification using whole-exome
- PDF / 561,294 Bytes
- 12 Pages / 595.276 x 785.197 pts Page_size
- 82 Downloads / 204 Views
RESEARCH ARTICLE Performance-weighted-voting model: an ensemble machine learning method for cancer type classification using whole-exome sequencing mutation Yawei Li, Yuan Luo* Department of Preventive Medicine, Northwestern University, Feinberg School of Medicine, Chicago, IL 60611, USA * Correspondence: [email protected] Received July 15, 2020; Revised August 31, 2020; Accepted September 2, 2020 Background: With improvements in next-generation DNA sequencing technology, lower cost is needed to collect genetic data. More machine learning techniques can be used to help with cancer analysis and diagnosis. Methods: We developed an ensemble machine learning system named performance-weighted-voting model for cancer type classification in 6,249 samples across 14 cancer types. Our ensemble system consists of five weak classifiers (logistic regression, SVM, random forest, XGBoost and neural networks). We first used cross-validation to get the predicted results for the five classifiers. The weights of the five weak classifiers can be obtained based on their predictive performance by solving linear regression functions. The final predicted probability of the performanceweighted-voting model for a cancer type can be determined by the summation of each classifier’s weight multiplied by its predicted probability. Results: Using the somatic mutation count of each gene as the input feature, the overall accuracy of the performanceweighted-voting model reached 71.46%, which was significantly higher than the five weak classifiers and two other ensemble models: the hard-voting model and the soft-voting model. In addition, by analyzing the predictive pattern of the performance-weighted-voting model, we found that in most cancer types, higher tumor mutational burden can improve overall accuracy. Conclusion: This study has important clinical significance for identifying the origin of cancer, especially for those where the primary cannot be determined. In addition, our model presents a good strategy for using ensemble systems for cancer type classification.
Keywords: cancer type classification; ensemble method; performance-weighted-voting model; linear regression; singlenucleotide polymorphism
Author summary: The identification of the cancer of unknown primary is of clinical significance, and can provide important cancer behavioral therapeutic strategies. To achieve this, we developed an ensemble machine learning system called the performance-weighted-voting model for cancer type classification. The ensemble system can integrate weak classifiers and train the weights of the weak classifiers based on their predictive performance. The model has achieved the highest overall accuracy among the models mentioned in this study. Furthermore, the model can theoretically promote any combination of weak classifiers with a high degree of accuracy.
INTRODUCTION It is generally accepted that tumorigenesis is a process of
cell renewal, replacement and accumulation of a series of oncogenes, tumor suppressor genes and genetic instability [1], resulting in the coll
Data Loading...
