引用本文
  •    [点击复制]
  •    [点击复制]
【打印本页】 【在线阅读全文】【下载PDF全文】 查看/发表评论下载PDF阅读器关闭

过刊浏览    高级检索

本文已被:浏览 166次   下载 0  
基于SNP标记的玉米自交系类群划分方法和分群功效评估指标的比较
李念念1, 王义波2, 徐国平2, 易黎2, 王爱方2, 李婷2, 曹刚强1
0
(1.郑州大学农学院;2.北京联创种业有限公司)
摘要:
随着高通量测序技术的不断进步带来的SNP 标记成本的持续下降,可用于种质资源分群和分子育种应用中基因型鉴定的SNP位点数越来越多,亟需系统地比较不同分群方法以便找到最合适的分群法和最可靠的分群功效评估指标。 本研究比较了4个分群方法(包括目前常用的邻接算法(NJ法)、SNPhylo法、ADMIXTURE + SNPs 和在ADMIXTURE +SNPs基础上开发的ADMIXTURE+TagSNPs分群法),以及4个分群功效评估指标(PCA散点图、群体遗传指标GD和PIC及贝叶斯统计指标BIC)在利用GBS简化基因组测序产生的525,141个SNPs位点数据将490份玉米自交系划分成3个和6个亚群时的表现。 结果表明:4个评估指标中的PCA散点图和BIC指标(BICBW,SBIC)探测亚群间变异的能力强,是评估不同分群方法分群功效的可靠指标,而GD和 PIC探测亚群间变异的能力差,不适合用作分群功效的评估。结果还表明,4个分群法均为有效分群法,所以都可用于种质资源分群,但ADMIXTURE+TagSNPs分群法划分的亚群边界清晰,亚群间个体混杂少,相对群间变异度大,综合表现最好,而SNPhylo法的综合表现最差。考虑到ADMIXTURE+TagSNPs需要输入的SNP标记数显著少于其他3种分群法,因而实际应用中基因型鉴定的成本最低,所以建议在遗传资源研究和分子育种应用中首选该分群法。
关键词:  玉米  聚类分析  标签SNP  主成分分析  遗传多样性  多态信息量  贝叶斯信息度
DOI:10.13430/j.cnki.jpgr.20190618003
投稿时间:2019-06-18修订日期:2020-02-16
基金项目:北京市科技计划重大项目 D171105007700003;河南省高等学校重点科研项目13A180687。
Comparison of Different Grouping Procedures and Evaluation Criteria for Grouping Maize Inbreds Using SNP Data
LI Nian-nian1, WANG Yi-bo2, SHU Guoping2, YI Li2, Wang Ai-fang2, Li Ting2, CAO Gang-qiang1
(1.School of Agricultural Sciences, Zhengzhou University;2.Beijing Lantron Seed Corporation)
Abstract:
Grouping germplasm lines and assisting plant breeding using large number of SNP marker have become well accepted due to constant price drops of SNP markers brought about by the advance at high-throughput sequencing technology. How to handle the large SNP datasets becomes an increasing interest, and the user-friendly statistical methodologies are in demand. In this study, four grouping procedures (NJ, SNPhylo , ADMIXTURE + SNPs, and the ADMIXTURE + TagSNPs which we modified from ADMIXTURE + SNPs), were deployed to group 490 corn inbreds into 3 and 6 subgroups using 525,141 SNP markers and their performance were evaluated with four criteria (PCA Scatter Plot, GD, PIC, and BIC). The result showed that PCA Scatter Plot and BIC (BICBW, SBIC)among the four criteria are more powerful in revealing between-subgroup variation, whereas GD and PIC showed less powerful. All four grouping procedures were effective and could be adopted in grouping germplasm. Particularly, ADMIXTURE+TagSNPs was the most effective in delineating subgroups with clear boundary and very little between-group mixing, while SNPhylo was the least effective. ADMIXTURE + TagSNPs required fewer SNP markers thus would cost less than other three procedures, and therefore was highly recommended for germplasm study and marker-assisted breeding.
Key words:  m aize  clustering analysis  tagSNP  principal component analysis  genetic diversity  polymorphism information content  Bayesian information criterion

用微信扫一扫

用微信扫一扫