Supplementary Material
SupplementaryFigure1. Heat map of the absolute values of D’ between the 26 SNPs in the study.
Supplementary Table 1.Results of 17 SNPs selected fromprevious GWASreported byYue et al9
Chr. / SNP ID / GWAS / Replication study / Combined analysis(746 cases,1,599 controls) / (4,027 cases, 5,603 controls)
MAF / MAF / PGWAS / OR / MAF / MAF / PReplication / OR / PCombined / OR
Case / Control / Case / Control
1 / rs1801133 / 0.423 / 0.488 / 0.00003 / 0.77 / 0.437 / 0.469 / 0.0004 / 0.88 / 0.005 / 0.83
3 / rs1381094 / 0.232 / 0.288 / 5.32E-04 / 3.12 / 0.086 / 0.084 / 0.771 / 1.02 / 0.302 / 1.78
4 / rs3111810 / 0.132 / 0.082 / 1.10E-07 / 1.7 / 0.148 / 0.108 / 0.0004 / 1.41 / 0.002 / 1.44
6 / rs1233710 / 0.259 / 0.326 / 4.14E-06 / 0.73 / 0.275 / 0.321 / 4.09E-07 / 0.8 / 4.76E-11 / 0.79
6 / rs1635 / 0.263 / 0.33 / 4.15E-06 / 0.73 / 0.275 / 0.325 / 5.53E-08 / 0.79 / 6.91E-12 / 0.78
6 / rs2142731 / 0.186 / 0.241 / 2.19E-05 / 0.72 / 0.184 / 0.22 / 9.15E-07 / 0.8 / 5.14E-10 / 0.79
7 / rs6978425 / 0.421 / 0.491 / 6.94E-06 / 0.75 / 0.316 / 0.354 / 8.96E-04 / 0.84 / 0.0007 / 0.81
9 / rs2065412 / 0.093 / 0.061 / 0.00005 / 1.59 / 0.075 / 0.064 / 0.02 / 1.18 / 0.041 / 1.35
11 / rs11038167 / 0.481 / 0.399 / 1.12E-07 / 1.4 / 0.453 / 0.401 / 3.28E-06 / 1.27 / 1.09E-11 / 1.29
11 / rs11038172 / 0.497 / 0.424 / 2.57E-06 / 1.34 / 0.47 / 0.419 / 1.11E-05 / 1.23 / 7.21E-10 / 1.25
11 / rs835784 / 0.343 / 0.277 / 4.02E-06 / 1.36 / 0.314 / 0.268 / 2.37E-07 / 1.25 / 2.73E-11 / 1.27
12 / rs3782886 / 0.158 / 0.205 / 0.00009 / 0.72 / 0.134 / 0.15 / 0.011 / 0.88 / 0.022 / 0.8
17 / rs11077721 / 0.28 / 0.227 / 0.00008 / 1.32 / 0.353 / 0.383 / 0.0008 / 0.88 / 0.008 / 0.82
17 / rs3744165 / 0.105 / 0.068 / 0.00002 / 1.59 / 0.086 / 0.077 / 0.069 / 1.13 / 0.104 / 1.32
17 / rs8073471 / 0.101 / 0.064 / 7.87E-06 / 1.64 / 0.08 / 0.072 / 0.089 / 1.12 / 0.118 / 1.35
18 / rs4939924 / 0.402 / 0.331 / 2.17E-06 / 1.36 / 0.389 / 0.368 / 0.02 / 1.09 / 0.082 / 1.21
20 / rs656111 / 0.389 / 0.315 / 8.62E-07 / 1.39 / 0.312 / 0.293 / 0.02 / 1.1 / 0.083 / 1.23
Abbreviations: Chr., chromosome; SNP, single-nucleotide polymorphism; MAF, minor allele frequency; OR, odds ratio; NA, not available.
Supplementary Table 2. Results of 9 SNPs selected from previous GWAS reported by Shi et al10
Chr. / SNP ID / GWAS / Replication / Combined analysis(3,750 cases, 6,468 controls) / (4,383 cases, 4,539 controls)
PGWAS / OR / PReplication / OR / PCombined
1 / rs10489202 / 2.65E-06 / 1.27 / 4.76E-05 / 1.19 / 9.50E-09
1 / rs1060041 / 4.11E-06 / 1.26 / 0.012 / 1.11 / 5.31E-07
1 / rs11586522 / 1.53E-06 / 1.25 / 0.362 / 1.04 / 1.17E-04
6 / rs7749823 / 4.18E-05 / 0.62 / NA / NA / NA
6 / rs6927023 / 2.93E-05 / 0.67 / NA / NA / NA
6 / rs2394514 / 1.31E-06 / 1.39 / NA / NA / NA
6 / rs1736913 / 4.60E-05 / 0.8 / NA / NA / NA
8 / rs16887244 / 9.27E-06 / 0.84 / 2.23E-08 / 0.83 / 1.27E-10
8 / rs1488935 / 2.81E-06 / 0.83 / 3.42E-05 / 0.87 / 5.06E-09
Abbreviations: Chr., chromosome; SNP, single-nucleotide polymorphism; OR, odds ratio; NA, not available.
SupplementaryTable 3. Gene expression level by all possible combination of rs3744165 and rs8073471Genotype / N / Expression / s.e. / CI
Cortex: TEAD3 (1.873E-8, FDR:1.710E-3)
AA or CA | AA / 1 / 6.179 / 0.242 / 5.653-6.706
AA or CA | GG or AG / 4 / 2.475 / 0.121 / 2.212-2.738
CC | AA / 10 / 2.460 / 0.076 / 2.294-2.626
Cortex: SH3TC2 (1.998E-8, FDR:1.170E-3)
AA or CA | AA / 1 / 4.649 / 0.166 / 4.287-5.012
AA or CA | GG or AG / 4 / 2.347 / 0.083 / 2.165-2.528
CC | AA / 10 / 2.074 / 0.053 / 1.959-2.188
Cortex: KCNK9 (5.195E-7, FDR:2.560E-2)
AA or CA | AA / 1 / 6.105 / 0.279 / 5.496-6.713
AA or CA | GG or AG / 4 / 2.942 / 0.140 / 2.638-3.246
CC | AA / 10 / 2.906 / 0.088 / 2.714-3.098
Cortex: PPDPF (1.132E-6, FDR: 4.780E-2)
AA or CA | AA / 1 / 6.713 / 0.369 / 5.909-7.517
AA or CA | GG or AG / 4 / 2.983 / 0.185 / 2.581-3.385
CC | AA / 10 / 2.743 / 0.117 / 2.489-2.998
Cerebellum: EFNA1 (7.255E-9,FDR: 1.710E-3)
AA or CA | AA / 1 / 6.937 / 0.310 / 6.268-7.606
AA or CA | GG or AG / 4 / 2.392 / 0.155 / 2.058-2.727
CC | AA / 11 / 2.156 / 0.093 / 1.954-2.358
Cerebellum: RNU4ATAC13P(2.318E-8, FDR: 1.710E-3)
AA or CA | AA / 1 / 5.939 / 0.312 / 5.265-6.613
AA or CA | GG or AG / 4 / 2.013 / 0.156 / 1.676-2.35
CC| AA / 11 / 1.555 / 0.094 / 1.352-1.758
Cerebellum: NUPL2 (6.788E-8, FDR: 4.010E-2)
AA or CA | AA / 1 / 5.223 / 0.201 / 4.789-5.656
AA or CA | GG or AG / 4 / 2.512 / 0.100 / 2.295-2.728
CC| AA / 11 / 2.796 / 0.061 / 2.665-2.926
Abbreviations and notes: s.e. Standard Error; The number in the bracket after the brain region and gene name is the P-value of the Ftest for joint significance.
Supplementary Information
Statistical methods for Case-control epistasis analysis:
We used logistic regression modelto test both additive and dominant genetic effects and adjust for age and BMI.
First the SNPs were tested one at a time to check the significance of their main effect. The model being used is:
where denote the probability for Observation i developing Schizophrenia, denote the intercept, is the age of Observation i, is the BMI of Observation i, is the square of BMI of Observation i, is the additive effect of SNP j and is the dominant effect deviated from additive effect of SNP j. , i=1 to 5, is the coeffcient of those variables.
Then logistic regression model was used to test pairwise gene-gene interaction for all combinations of the 26 SNPs. Two models were fitted as follows,
where denote the probability for Observation i developing Schizophrenia, denote the intercept, is the age of Observation i, is the BMI of Observation i, is the square of BMI of Observation i, is the additive effect of SNP j and is the dominant effect deviated from additive effect of SNP j, is the additive effect of SNP k and is the dominant effect deviated from additive effect of SNP k. , i=1 to 11, is the coeffcient of those variables.
Deviance was calculated for both models and test was performed to test the over all significance of all the interaction effect.
Statistical methods for eQTLepistasis analysis:
We tested for association between gene expression level and haplotype of the two spotted SNPs using a linear model approach.
For those sub-regions in which each subject only has one sample, one linear model was fitted as follows.
where denote the intercept, is an indicator variable which denotes whether the sample is a certain haplotype and is the coeffcient of . F test was used to test the overall significance of all . False Discovery Rate (FDR) corrections were applied to control the family wise error rate.
For those sub-regions in which each subject only has multiple samples, two linear model was fitted as follows.
where and have the same meaning as the previous one, is a random effect which accounts for the variability brought by the subject. Deviance test was used to test the overall significance of all . Again, False Discovery Rate (FDR) corrections were applied to control the family wise error rate as well.
Genetic REsearch on schizophreniAneTwork-China and Netherland (GREAT-CN) membership
Lili Guan1, Baoming Wu1, Hong Ma1, Xin Yu1, Zhida Xu2, Rene Kahn2, Min Zhang1,3, Dabao Zhang3, Qi Wang3, Libo Wang3, Yuantang Chen4, Junyan Zhu4, Haiyan Zhang4, Shan Wang4, Chao Li4, Feihu Liu4, Jianguo Shi4, Tao Zhang5, Fenghua Ye5, Xingrong Wu5, Lihong Jia5, Yongqi Shi5, Baoping Yan6, Yongqiao Liu6, Lijun Cui6, Dongrui Zheng6, Yunshu Zhang6, Keqing Li6, Lei Su7, Xiaodong Yu7, Chaojuan Mu7, Jia Jian7, Guolin Mi7, Lanfen Liu7, Aizhen Wang7, Songbo Wang7, Tianliang Zhang7, Dongdong Qiao7, Chuanhua Lu7, Hongxia Zhang8, Guinan Jin8, Xiangqiao Xiao8, Yuyan Shi8, Hongli Wang8, Baode Yu8, Zhiming Zhang8, Rongyi Li8, Jun Hong8, Yan Song8, Rui Jiang9, Lili Wang9, Jiang Cao9, Xiaoju Jia9, Conghui Wang9, Tian Tian9, Shuqin Zhang, 9Changjun Du9, Guifu Yang9, Hongjun Tian9, Yiquan Wang10, Defeng Hu10, Zhiyu Chen10, Ziming Xu10, Yajun Jin10, Yunhai Tao10, Yali Mo10, Wan Ma10, Yi Cheng10, Junshan Chen11, Yuan Yuan11, Bei Wei11, Qingjun Li11, Daqi Li11,Yujuan Ma11, Guojian Xie11, Jie Ning11, Bo Wang11, Renchun Huang12, Mingjian Zheng12, Saizheng Weng12, Wenqian Ou12, Hong Deng12, Xiaodan Li12, Weidong Cong12, Jingtao Zheng12, Jicai Wang13, Linling Jiang13, Xiufeng Xu13, Hao Liu14, Can Yang14, Weidong Xiao14, Xiaoping Wang14, Gaohua Wang14, Chao Han15, Yushi Zhang15, Jun Shao15, Shuo Li15, Xingju Qiao16, Xiangxin Liu16, Fujun Jia16
1Peking University Sixth Hospital, Peking University Institute of Mental Health, Key Laboratory of Mental Health, Ministry of Health (Peking University), Beijing, China
2Division of Neuroscience of the University Medical Centre in Utrecht, Utrecht, the Netherlands
3Department of Statistics, Purdue University, West Lafayette, USA
4Xi'an Mental Health Center, Xi’an, China
5Shanxi Provincial Mental Health Center, Taiyuan, China
6The Mental Health Center of Hebei Province, Baoding, China
7Shandong Mental Health Center, Jinan, China
8Mudanjiang Neuro-Psychiatric Hospital of Heilongjiang Province, Mudanjiang, China
9Tianjin Anding Hospital, Tianjin, China
10Hangzhou Seventh People's Hospital, Hangzhou, China
11Chongqing Mental Health Center, Chongqing, China
12Neuro-Psychiatric Hospital of Fuzhou, Fujian Province, Fuzhou, China
13The First Affiliated Hospital of Kunming Medical University, Kunming, China
14Renmin Hospital of Wuhan University, Wuhan, China
15Xicheng Ping’an Hospital, Beijing, China
16Institute of Guangdong Province, Guangzhou, China
1