Permutation-based significance analysis reduces the type 1 error rate in bisulphite sequencing data analysis of human umbilical cord blood samples. Issue 12 (2nd December 2022)

• Record Type: Journal Article
• Authors: Laajala, Essi; Halla-aho, Viivi; Grönroos, Toni; Kalim, Ubaid Ullah; Vähä-Mäkilä, Mari; Nurmio, Mirja; Kallionpää, Henna; Lietzén, Niina; Mykkänen, Juha; Rasool, Omid; Toppari, Jorma; Orešič, Matej; Knip, Mikael; Lund, Riikka; Lahesmaa, Riitta; Lähdesmäki, Harri
• Is Part Of: Epigenetics. Volume 17:Issue 12(2022)
• Page Start: 1608
• Page End: 1627
• Publication Date: 2022-12-02
• Subjects: DNA methylation -- bisulphite sequencing -- RRBS -- umbilical cord blood -- pregnancy -- sex -- spatial correlation -- type 1 error -- differential methylation -- analysis workflow; Epigenesis -- Periodicals; Epigenetica; 572.86505
• DOI: 10.1080/15592294.2022.2044127
• Languages: English
• ISSNs: 1559-2294

Abstract

ABSTRACT: DNA methylation patterns are largely established in-utero and might mediate the impacts of in-utero conditions on later health outcomes. Associations between perinatal DNA methylation marks and pregnancy-related variables, such as maternal age and gestational weight gain, have been earlier studied with methylation microarrays, which typically cover less than 2% of human CpG sites. To detect such associations outside these regions, we chose the bisulphite sequencing approach. We collected and curated clinical data on 200 newborn infants; whose umbilical cord blood samples were analysed with the reduced representation bisulphite sequencing (RRBS) method. A generalized linear mixed-effects model was fit for each high coverage CpG site, followed by spatial and multiple testing adjustment of P values to identify differentially methylated cytosines (DMCs) and regions (DMRs) associated with clinical variables, such as maternal age, mode of delivery, and birth weight. Type 1 error rate was then evaluated with a permutation analysis. We discovered a strong inflation of spatially adjusted P values through the permutation analysis, which we then applied for empirical type 1 error control. The inflation of P values was caused by a common method for spatial adjustment and DMR detection, implemented in tools comb-p and RADMeth. Based on empirically estimated significance thresholds, very little differential methylation was associated with any of the studied clinical variables, other than sex. With this analysis workflow, the sex-associated differentially methylated regions were highly reproducible across studies, technologies, and statistical models.