Powerful p-value combination methods to detect incomplete association.
Powerful p-value combination methods to detect incomplete association.
复制标题
DOI：
10.1038/s41598-021-86465-y
复制
发表时间：
2021-03-26
影响因子：
4.6
通讯作者：
Nam D
中科院分区：
文献类型：
--
作者：
Yoon S;Baik B;Park T;Nam D
荟萃分析通过结合多项研究的统计数据来增加统计功效。荟萃分析方法大多是在每个研究中的所有数据都与给定表型相关的条件下进行评估的。然而，每个研究中的特定实验条件或遗传异质性可能导致来自零分布的“不相关统计量”。在这里，我们表明，传统的荟萃分析方法的功率迅速下降，包括越来越多的无关联的统计数据，而经典的Fisher方法及其加权的变体（wFisher）表现出相对较高的功率，是强大的无关联的统计数据。我们还提出了另一种基于有序p值联合分布（ordmeta）的鲁棒方法。对t检验、RNA-seq和微阵列数据的模拟分析表明，当只有少数研究存在关联时，wFisher和ordmeta优于现有的荟萃分析方法。我们对9个微阵列数据集（前列腺癌）和4个关联汇总数据集（体重指数）进行了荟萃分析，其中我们的方法表现出高度的生物相关性，并且能够检测到最先进方法遗漏的基因。实现所提出方法的metapro R包可从CRAN和GitHub（http：//github.com/unistbig/metapro）获得。
Meta-analyses increase statistical power by combining statistics from multiple studies. Meta-analysis methods have mostly been evaluated under the condition that all the data in each study have an association with the given phenotype. However, specific experimental conditions in each study or genetic heterogeneity can result in “unassociated statistics” that are derived from the null distribution. Here, we show that power of conventional meta-analysis methods rapidly decreases as an increasing number of unassociated statistics are included, whereas the classical Fisher’s method and its weighted variant (wFisher) exhibit relatively high power that is robust to addition of unassociated statistics. We also propose another robust method based on joint distribution of ordered p-values (ordmeta). Simulation analyses for t-test, RNA-seq, and microarray data demonstrated that wFisher and ordmeta, when only a small number of studies have an association, outperformed existing meta-analysis methods. We performed meta-analyses of nine microarray datasets (prostate cancer) and four association summary datasets (body mass index), where our methods exhibited high biological relevance and were able to detect genes that the-state-of-the-art methods missed. The metapro R package that implements the proposed methods is available from both CRAN and GitHub (http://github.com/unistbig/metapro).
登录
查看更多内容
影响因子：
30.8
作者：
DIAbetes Genetics Replication And Meta-analysis (DIAGRAM) Consortium;Asian Genetic Epidemiology Network Type 2 Diabetes (AGEN-T2D) Consortium;South Asian Type 2 Diabetes (SAT2D) Consortium;Mexican American Type 2 Diabetes (MAT2D) Consortium;Type 2 Diabetes Genetic Exploration by Nex-generation sequencing in muylti-Ethnic Samples (T2D-GENES) Consortium;Mahajan A;Go MJ;Zhang W;Below JE;Gaulton KJ;Ferreira T;Horikoshi M;Johnson AD;Ng MC;Prokopenko I;Saleheen D;Wang X;Zeggini E;Abecasis GR;Adair LS;Almgren P;Atalay M;Aung T;Baldassarre D;Balkau B;Bao Y;Barnett AH;Barroso I;Basit A;Been LF;Beilby J;Bell GI;Benediktsson R;Bergman RN;Boehm BO;Boerwinkle E;Bonnycastle LL;Burtt N;Cai Q;Campbell H;Carey J;Cauchi S;Caulfield M;Chan JC;Chang LC;Chang TJ;Chang YC;Charpentier G;Chen CH;Chen H;Chen YT;Chia KS;Chidambaram M;Chines PS;Cho NH;Cho YM;Chuang LM;Collins FS;Cornelis MC;Couper DJ;Crenshaw AT;van Dam RM;Danesh J;Das D;de Faire U;Dedoussis G;Deloukas P;Dimas AS;Dina C;Doney AS;Donnelly PJ;Dorkhan M;van Duijn C;Dupuis J;Edkins S;Elliott P;Emilsson V;Erbel R;Eriksson JG;Escobedo J;Esko T;Eury E;Florez JC;Fontanillas P;Forouhi NG;Forsen T;Fox C;Fraser RM;Frayling TM;Froguel P;Frossard P;Gao Y;Gertow K;Gieger C;Gigante B;Grallert H;Grant GB;Grrop LC;Groves CJ;Grundberg E;Guiducci C;Hamsten A;Han BG;Hara K;Hassanali N;Hattersley AT;Hayward C;Hedman AK;Herder C;Hofman A;Holmen OL;Hovingh K;Hreidarsson AB;Hu C;Hu FB;Hui J;Humphries SE;Hunt SE;Hunter DJ;Hveem K;Hydrie ZI;Ikegami H;Illig T;Ingelsson E;Islam M;Isomaa B;Jackson AU;Jafar T;James A;Jia W;Jöckel KH;Jonsson A;Jowett JB;Kadowaki T;Kang HM;Kanoni S;Kao WH;Kathiresan S;Kato N;Katulanda P;Keinanen-Kiukaanniemi KM;Kelly AM;Khan H;Khaw KT;Khor CC;Kim HL;Kim S;Kim YJ;Kinnunen L;Klopp N;Kong A;Korpi-Hyövälti E;Kowlessur S;Kraft P;Kravic J;Kristensen MM;Krithika S;Kumar A;Kumate J;Kuusisto J;Kwak SH;Laakso M;Lagou V;Lakka TA;Langenberg C;Langford C;Lawrence R;Leander K;Lee JM;Lee NR;Li M;Li X;Li Y;Liang J;Liju S;Lim WY;Lind L;Lindgren CM;Lindholm E;Liu CT;Liu JJ;Lobbens S;Long J;Loos RJ;Lu W;Luan J;Lyssenko V;Ma RC;Maeda S;Mägi R;Männisto S;Matthews DR;Meigs JB;Melander O;Metspalu A;Meyer J;Mirza G;Mihailov E;Moebus S;Mohan V;Mohlke KL;Morris AD;Mühleisen TW;Müller-Nurasyid M;Musk B;Nakamura J;Nakashima E;Navarro P;Ng PK;Nica AC;Nilsson PM;Njølstad I;Nöthen MM;Ohnaka K;Ong TH;Owen KR;Palmer CN;Pankow JS;Park KS;Parkin M;Pechlivanis S;Pedersen NL;Peltonen L;Perry JR;Peters A;Pinidiyapathirage JM;Platou CG;Potter S;Price JF;Qi L;Radha V;Rallidis L;Rasheed A;Rathman W;Rauramaa R;Raychaudhuri S;Rayner NW;Rees SD;Rehnberg E;Ripatti S;Robertson N;Roden M;Rossin EJ;Rudan I;Rybin D;Saaristo TE;Salomaa V;Saltevo J;Samuel M;Sanghera DK;Saramies J;Scott J;Scott LJ;Scott RA;Segrè AV;Sehmi J;Sennblad B;Shah N;Shah S;Shera AS;Shu XO;Shuldiner AR;Sigurđsson G;Sijbrands E;Silveira A;Sim X;Sivapalaratnam S;Small KS;So WY;Stančáková A;Stefansson K;Steinbach G;Steinthorsdottir V;Stirrups K;Strawbridge RJ;Stringham HM;Sun Q;Suo C;Syvänen AC;Takayanagi R;Takeuchi F;Tay WT;Teslovich TM;Thorand B;Thorleifsson G;Thorsteinsdottir U;Tikkanen E;Trakalo J;Tremoli E;Trip MD;Tsai FJ;Tuomi T;Tuomilehto J;Uitterlinden AG;Valladares-Salgado A;Vedantam S;Veglia F;Voight BF;Wang C;Wareham NJ;Wennauer R;Wickremasinghe AR;Wilsgaard T;Wilson JF;Wiltshire S;Winckler W;Wong TY;Wood AR;Wu JY;Wu Y;Yamamoto K;Yamauchi T;Yang M;Yengo L;Yokota M;Young R;Zabaneh D;Zhang F;Zhang R;Zheng W;Zimmet PZ;Altshuler D;Bowden DW;Cho YS;Cox NJ;Cruz M;Hanis CL;Kooner J;Lee JY;Seielstad M;Teo YY;Boehnke M;Parra EJ;Chambers JC;Tai ES;McCarthy MI;Morris AP
通讯作者：
Morris AP
影响因子：
14.9
作者：
Piñero J;Bravo À;Queralt-Rosinach N;Gutiérrez-Sacristán A;Deu-Pons J;Centeno E;García-García J;Sanz F;Furlong LI
通讯作者：
Furlong LI
影响因子：
8
作者：
Plymate, SR;Haugk, KH;Oberley, TD
通讯作者：
Oberley, TD
影响因子：
3
作者：
Mägi R;Morris AP
通讯作者：
Morris AP
影响因子：
2.2
作者：
DerSimonian R;Laird N
通讯作者：
Laird N