Quantify and control reproducibility in high-throughput experiments.
Quantify and control reproducibility in high-throughput experiments.
复制标题
DOI:
10.1038/s41592-020-00978-4
复制
发表时间:
2020-12
期刊:
影响因子:
48
通讯作者:
Wen X
中科院分区:
文献类型:
--
作者:
Zhao Y;Sampson MG;Wen X
We propose a set of computational methods, named INTRIGUE, to evaluate and control reproducibility in high-throughput experiments. Our approaches are built upon a novel definition of reproducibility, which emphasizes directional consistency (DC) when experimental units are assessed with signed effect size estimates. The proposed methods are designed to i) assess the overall reproducible quality of multiple studies; and ii) evaluate reproducibility at the level of individual experimental units. We demonstrate the proposed methods in detecting unobserved batch effects in high-throughput experiments via simulations. In an application of assessing reproducibility in transcriptome-wide association studies (TWAS), we illustrate the versatility of the proposed methods: in addition to reproducible quality control, they are also suited for investigating genuine biological heterogeneity. Finally, we discuss the extensions of the proposed reproducibility measures and potential applications in other vital areas of reproducible research (e.g., publication bias and conceptual replications).
登录
查看更多内容
影响因子:
2.1
作者:
Li, Gen;Shabalin, Andrey A.;Nobel, Andrew B.
通讯作者:
Nobel, Andrew B.
影响因子:
7
作者:
Augusteijn, Hilde E. M.;van Aert, Robbie C. M.;van Assen, Marcel A. L. M.
通讯作者:
van Assen, Marcel A. L. M.
DOI:
10.1111/j.1467-985x.2009.00629.x
发表时间:
2010-01-01
影响因子:
2
作者:
Peters, Jaime L.;Sutton, Alex J.;Moreno, Santiago G.
通讯作者:
Moreno, Santiago G.
影响因子:
17.1
作者:
Goodman, Steven N.;Fanelli, Daniele;Ioannidis, John P. A.
通讯作者:
Ioannidis, John P. A.
影响因子:
4.2
作者:
Schmidt, Stefan
通讯作者:
Schmidt, Stefan