An application of Random Forests to a genome-wide association dataset: methodological considerations & new findings.
An application of Random Forests to a genome-wide association dataset: methodological considerations & new findings.
复制标题
DOI：
10.1186/1471-2156-11-49
复制
发表时间：
2010-06-14
期刊：
影响因子：
2.9
通讯作者：
Barcellos LF
中科院分区：
文献类型：
--
作者：
Goldstein BA;Hubbard AE;Cutler A;Barcellos LF
随着计算能力的提高，将更先进的机器学习技术应用于大型全基因组关联(GWA)数据集的分析成为可能。虽然大多数传统的统计方法只能阐明遗传变异对疾病风险的主要影响，但某些机器学习方法特别适合于发现高阶和非线性影响。一种这样的方法是随机森林(RF)算法。近年来，RF用于发现与人类疾病相关的SNP的情况有所增加；然而，大多数工作都集中在小数据集或模拟研究上，这些研究都是有限的。使用一个包含全基因组300K SNP基因型的多发性硬化症(MS)病例对照数据集，我们概述了一种基于经验数据集优化RF算法的方法和一些考虑因素。重要的是，结果表明典型的默认参数值不适合大型GWA数据集。此外，可以通过对数据进行二次采样、基于连锁不平衡(LD)的修剪以及去除RF分析中的强烈影响来获得收益。新的RF结果与最初的MS GWA研究结果进行了比较，显示出重叠。此外，通过RF分析确定了四个新的候选MS基因，MPHOSPH9、CTNNA3、PHACTR2和IL7，并需要在独立研究中进行进一步的后续研究。这项研究是用机器学习算法成功分析GWA数据的第一批例证之一。结果表明，对于GWA数据，RF在计算上是可行的，在前人研究的基础上所得到的结果具有生物学意义。更重要的是，新的基因被确定为可能与多发性硬化症相关，这为这种复杂疾病的研究提供了新的途径。
As computational power improves, the application of more advanced machine learning techniques to the analysis of large genome-wide association (GWA) datasets becomes possible. While most traditional statistical methods can only elucidate main effects of genetic variants on risk for disease, certain machine learning approaches are particularly suited to discover higher order and non-linear effects. One such approach is the Random Forests (RF) algorithm. The use of RF for SNP discovery related to human disease has grown in recent years; however, most work has focused on small datasets or simulation studies which are limited. Using a multiple sclerosis (MS) case-control dataset comprised of 300 K SNP genotypes across the genome, we outline an approach and some considerations for optimally tuning the RF algorithm based on the empirical dataset. Importantly, results show that typical default parameter values are not appropriate for large GWA datasets. Furthermore, gains can be made by sub-sampling the data, pruning based on linkage disequilibrium (LD), and removing strong effects from RF analyses. The new RF results are compared to findings from the original MS GWA study and demonstrate overlap. In addition, four new interesting candidate MS genes are identified, MPHOSPH9, CTNNA3, PHACTR2 and IL7, by RF analysis and warrant further follow-up in independent studies. This study presents one of the first illustrations of successfully analyzing GWA data with a machine learning algorithm. It is shown that RF is computationally feasible for GWA data and the results obtained make biologic sense based on previous studies. More importantly, new genes were identified as potentially being associated with MS, suggesting new avenues of investigation for this complex disease.
登录
查看更多内容
影响因子：
2.5
作者：
Wider C;Lincoln SJ;Heckman MG;Diehl NN;Stone JT;Haugarvoll K;Aasly JO;Gibson JM;Lynch T;Rajput A;Rajput ML;Uitti RJ;Wszolek ZK;Farrer MJ;Ross OA
通讯作者：
Ross OA
影响因子：
120.7
作者：
Pearson, Thomas A.;Manolio, Teri A.
通讯作者：
Manolio, Teri A.
随机森林中的偏见重要性措施：插图，来源和解决方案。
影响因子：
3
作者：
Strobl, Carolin;Boulesteix, Anne-Laure;Zeileis, Achim;Hothorn, Torsten
通讯作者：
Hothorn, Torsten
影响因子：
3
作者：
Meng, Yan A.;Yu, Yi;Lunetta, Kathryn L.
通讯作者：
Lunetta, Kathryn L.
影响因子：
30.8
作者：
De Jager, Philip L.;Jia, Xiaoming;Wang, Joanne;de Bakker, Paul I. W.;Ottoboni, Linda;Aggarwal, Neelum T.;Piccio, Laura;Raychaudhuri, Soumya;Tran, Dong;Aubin, Cristin;Briskin, Rebeccah;Romano, Susan;Baranzini, Sergio E.;McCauley, Jacob L.;Pericak-Vance, Margaret A.;Haines, Jonathan L.;Gibson, Rachel A.;Naeglin, Yvonne;Uitdehaag, Bernard;Matthews, Paul M.;Kappos, Ludwig;Polman, Chris;McArdle, Wendy L.;Strachan, David P.;Evans, Denis;Cross, Anne H.;Daly, Mark J.;Compston, Alastair;Sawcer, Stephen J.;Weiner, Howard L.;Hauser, Stephen L.;Hafler, David A.;Oksenberg, Jorge R.
通讯作者：
Oksenberg, Jorge R.