摘要
对急性髓性白血病(AML)病人进行明确的亚型分类,有助于制定合适的治疗方案并预测其治疗效果。之前研究表明基因芯片技术在白血病亚型分类中已取得了较好效果,但由于儿童AML发病率较低,相应的芯片分析研究较少,因此目前用于构建儿童AML亚型分类模型的数据相对不足,是否可以应用现有的成人分类模型数据来对儿童AML进行预报还有待研究。应用基因芯片整合分析方法,对来自不同实验的研究成人或儿童AML亚型分类的基因芯片数据进行整合,应用支持向量机分析整合后数据集的亚型预报准确率。结果表明整合后的芯片数据在儿童AML亚型分类预报中的准确率达到97.24%,特征基因分析结果也说明在同一种AML亚型中,对于来自不同年龄组的样本,其特征基因有较高的表达相似性。
Accurate subgroup classification for Acute Myeloid Leukemia patient will be of great help in providing proper treatment. The extensive use of DNA microarray technology in the AML classification study has been proved good results. AML is a relatively rare malignancy in the pediatric population. The pediatric microarray data is not enough to build subgroup classification model with high accuracy. In this study, we applied microarray integration analysis on adults and pediatric AML gene expression data. Results showed that high classification accuracy 97.24% was obtained and same expression pattern was discovered within the same subgroup among different age groups, indicating that classification model with high accuracy can be set up for Pediatric AML samples.
出处
《生物信息学》
2009年第2期99-103,共5页
Chinese Journal of Bioinformatics
