吉林大学学报(理学版) ›› 2021, Vol. 59 ›› Issue (5): 1252-1255.

• • 上一篇    下一篇

一种基于集成学习策略的单细胞转录组数据集成分类算法

刘桂锋1, 于绍楠1, 崔璐2   

  1. 1. 吉林大学中日联谊医院 放射线科, 长春 130033; 2. 吉林大学中日联谊医院 医疗保险管理部, 长春 130033
  • 收稿日期:2021-05-18 出版日期:2021-09-26 发布日期:2021-09-26
  • 通讯作者: 崔璐 E-mail:cuilu@jlu.edu.cn

An Ensemble Classification Algorithm for Single Cell Transcriptome Data Based on Ensemble Learning Strategy

LIU Guifeng1, YU Shaonan1, CUI Lu2   

  1. 1. Department of Radiology, China-Japan Union Hospital of Jilin University, Changchun 130033, China;
    2. Department of Medical Insurance, China-Japan Union Hospital of Jilin University, Changchun 130033, China
  • Received:2021-05-18 Online:2021-09-26 Published:2021-09-26

摘要: 针对单细胞转录组数据上细胞分类准确率较低的问题, 提出一种新的细胞集成分类算法. 该方法能充分利用不同分类模型的优点, 降低单细胞数据的分类误差. 分别在慢性粒细胞白血病单细胞测序数据和三阴性乳腺癌单细胞测序数据两个不同数据集上进行实验验证, 实验结果表明, 由集成算法划分的细胞分类更清晰准确, 验证了该算法的有效性.

关键词: 单细胞转录组, 集成分类模型, k-近邻算法, 支持向量机

Abstract: Aiming at the problem of low accuracy of the cell classification of single cell transcriptome data, we proposed a novel cell ensemble classification algorithm. The algorithm could make full use of advantages of different classification models and reduce the classification error of single cell data. The experimental results on a chronic myeloid leukemia data and a triple-negative breast cancer data show that the cell classification based on the ensemble algorithm is more clear and accurate, which verifies the effectiveness of the proposed algorithm.

Key words: single cell transcriptome, ensemble classification model, k-nearest neighbor algorithm, support vector machine

中图分类号: 

  • TP751