10.6084/m9.figshare.c.3620078_D4.v1 Andrew Li Yim Andrew Li Yim Nicolette Duijvis Nicolette Duijvis Jing Zhao Jing Zhao Wouter de Jonge Wouter de Jonge Geert D’Haens Geert D’Haens Marcel Mannens Marcel Mannens Adri Mul Adri Mul Anje te Velde Anje te Velde Peter Henneman Peter Henneman Additional file 1: Figure S1. of Peripheral blood methylation profiling of female Crohn’s disease patients Springer Nature 2016 Crohn’s disease Inflammatory bowel diseases Females DNA methylation Peripheral blood Epigenome-wide association study 2016-06-08 05:00:00 Journal contribution https://springernature.figshare.com/articles/journal_contribution/Additional_file_1_Figure_S1_of_Peripheral_blood_methylation_profiling_of_female_Crohn_s_disease_patients/4391165 Exploratory data analysis of putative biological confounders. a) Variance explained per principal component based on our 450k data (turquoise triangles) versus randomly generated data (red circles). b) Dot-boxplot of the cellular composition as estimated by the Houseman algorithm [22, 23]. c) Pearson correlation coefficient (R2) of each blood cell proportion with the principal components. The statistically significant correlations are indicated in turquoise, whereas statistically non-significant associations are indicated in red. Similar correlations were calculated for d) age and e) anti-TNF usage. (PDF 119 kb)