Stability selection enhances feature selection and enables accurate prediction of gestational age using only five DNA methylation sites.
Stability selection enhances feature selection and enables accurate prediction of gestational age using only five DNA methylation sites.
复制标题
稳定性选择可增强特征选择，并仅使用五个DNA甲基化位点才能准确预测胎龄。
DOI：
10.1186/s13148-023-01528-3
复制
发表时间：
2023-07-13
影响因子：
5.7
通讯作者：
Denault, William R. P.
中科院分区：
文献类型：
--
作者：
Haftorn, Kristine L. L.;Romanowska, Julia;Lee, Yunsung;Page, Christian M. M.;Magnus, Per M. M.;Haberg, Siri E. E.;Bohlin, Jon;Jugessur, Astanand;Denault, William R. P.
关键词：
DNA 甲基化 (DNAm) 与儿童和成人的实际年龄以及新生儿的胎龄 (GA) 密切相关。这一特性使得多种表观遗传时钟得以开发，可以准确预测实际年龄和 GA。然而，不同表观遗传时钟之间的预测 CpG 缺乏重叠仍然难以捉摸。因此，我们的主要目标是识别和表征能够稳定预测 GA 的 CpG。 我们对挪威母亲、父亲和儿童队列研究中 2138 名新生儿的 DNAm 数据应用了一种称为“稳定性选择”的统计方法。稳定性选择将二次采样与变量选择相结合，以限制所选变量集中错误发现的数量。 24 个 CpG 被确定为可以稳定预测 GA。有趣的是，之前的 GA 时钟中只有 10% 的 CpG 被稳定选择。基于这些结果，我们使用广义加性模型回归开发了一种仅由 5 个 CpG 组成的新 GA 时钟，其显示出与之前的 GA 时钟相似的预测性能（R2 = 0.674，中位绝对偏差 = 4.4 天）。这些 CpG 位于或靠近参与免疫反应、代谢和发育过程的基因和调节区域。此外，考虑非线性关联可以提高早产儿的预测性能。我们提出了一个特征选择的方法框架，该框架广泛适用于可以从 DNAm 数据预测的任何性状。我们通过识别对 GA 具有高度预测性的 CpG 来展示其实用性，并提出一种仅基于 5 个更适合临床环境的 CpG 的新型高性能 GA 时钟。在线版本包含可在 10.1186/s13148-023-01528-3 获取的补充材料。
DNA methylation (DNAm) is robustly associated with chronological age in children and adults, and gestational age (GA) in newborns. This property has enabled the development of several epigenetic clocks that can accurately predict chronological age and GA. However, the lack of overlap in predictive CpGs across different epigenetic clocks remains elusive. Our main aim was therefore to identify and characterize CpGs that are stably predictive of GA. We applied a statistical approach called ‘stability selection’ to DNAm data from 2138 newborns in the Norwegian Mother, Father, and Child Cohort study. Stability selection combines subsampling with variable selection to restrict the number of false discoveries in the set of selected variables. Twenty-four CpGs were identified as being stably predictive of GA. Intriguingly, only up to 10% of the CpGs in previous GA clocks were found to be stably selected. Based on these results, we used generalized additive model regression to develop a new GA clock consisting of only five CpGs, which showed a similar predictive performance as previous GA clocks (R2 = 0.674, median absolute deviation = 4.4 days). These CpGs were in or near genes and regulatory regions involved in immune responses, metabolism, and developmental processes. Furthermore, accounting for nonlinear associations improved prediction performance in preterm newborns. We present a methodological framework for feature selection that is broadly applicable to any trait that can be predicted from DNAm data. We demonstrate its utility by identifying CpGs that are highly predictive of GA and present a new and highly performant GA clock based on only five CpGs that is more amenable to a clinical setting. The online version contains supplementary material available at 10.1186/s13148-023-01528-3.
登录
查看更多内容
影响因子：
5.7
作者：
Haftorn KL;Lee Y;Denault WRP;Page CM;Nustad HE;Lyle R;Gjessing HK;Malmberg A;Magnus MC;Næss Ø;Czamara D;Räikkönen K;Lahti J;Magnus P;Håberg SE;Jugessur A;Bohlin J
通讯作者：
Bohlin J
影响因子：
14.9
作者：
Howe KL;Achuthan P;Allen J;Allen J;Alvarez-Jarreta J;Amode MR;Armean IM;Azov AG;Bennett R;Bhai J;Billis K;Boddu S;Charkhchi M;Cummins C;Da Rin Fioretto L;Davidson C;Dodiya K;El Houdaigui B;Fatima R;Gall A;Garcia Giron C;Grego T;Guijarro-Clarke C;Haggerty L;Hemrom A;Hourlier T;Izuogu OG;Juettemann T;Kaikala V;Kay M;Lavidas I;Le T;Lemos D;Gonzalez Martinez J;Marugán JC;Maurel T;McMahon AC;Mohanan S;Moore B;Muffato M;Oheh DN;Paraschas D;Parker A;Parton A;Prosovetskaia I;Sakthivel MP;Salam AIA;Schmitt BM;Schuilenburg H;Sheppard D;Steed E;Szpak M;Szuba M;Taylor K;Thormann A;Threadgold G;Walts B;Winterbottom A;Chakiachvili M;Chaubal A;De Silva N;Flint B;Frankish A;Hunt SE;IIsley GR;Langridge N;Loveland JE;Martin FJ;Mudge JM;Morales J;Perry E;Ruffier M;Tate J;Thybert D;Trevanion SJ;Cunningham F;Yates AD;Zerbino DR;Flicek P
通讯作者：
Flicek P
DOI：
10.1093/bioinformatics/btx346
发表时间：
2017-10-01
期刊：
Bioinformatics (Oxford, England)
影响因子：
--
作者：
Gel B;Serra E
通讯作者：
Serra E
影响因子：
29
作者：
Barish GD;Yu RT;Karunasiri MS;Becerra D;Kim J;Tseng TW;Tai LJ;Leblanc M;Diehl C;Cerchietti L;Miller YI;Witztum JL;Melnick AM;Dent AL;Tangirala RK;Evans RM
通讯作者：
Evans RM
影响因子：
64.8
作者：
Jepsen, Kristen;Solum, Derek;Rosenfeld, Michael G.
通讯作者：
Rosenfeld, Michael G.