DNA methylation in peripheral tissues and left-handedness

Handedness has low heritability and epigenetic mechanisms have been proposed as an etiological mechanism. To examine this hypothesis, we performed an epigenome-wide association study of left-handedness. In a meta-analysis of 3914 adults of whole-blood DNA methylation, we observed that CpG sites located in proximity of handedness-associated genetic variants were more strongly associated with left-handedness than other CpG sites (P = 0.04), but did not identify any differentially methylated positions. In longitudinal analyses of DNA methylation in peripheral blood and buccal cells from children (N = 1737), we observed moderately stable associations across age (correlation range [0.355–0.578]), but inconsistent across tissues (correlation range [− 0.384 to 0.318]). We conclude that DNA methylation in peripheral tissues captures little of the variance in handedness. Future investigations should consider other more targeted sources of tissue, such as the brain.

Handedness has low heritability and epigenetic mechanisms have been proposed as an etiological mechanism. To examine this hypothesis, we performed an epigenome-wide association study of left-handedness. In a meta-analysis of 3914 adults of whole-blood DNA methylation, we observed that CpG sites located in proximity of handedness-associated genetic variants were more strongly associated with left-handedness than other CpG sites (P = 0.04), but did not identify any differentially methylated positions. In longitudinal analyses of DNA methylation in peripheral blood and buccal cells from children (N = 1737), we observed moderately stable associations across age (correlation range [0.355-0.578]), but inconsistent across tissues (correlation range [− 0.384 to 0.318]). We conclude that DNA methylation in peripheral tissues captures little of the variance in handedness. Future investigations should consider other more targeted sources of tissue, such as the brain.
Handedness, defined as the preferential use of one hand over the other, is established early in life and represents a highly stable trait that is thought to be accompanied by changes in brain 1 , corticospinal tract 2 , peripheral innervation and vascularization of arm skeletal muscles 3 , arm dynamics 4 , and possibly the immune system 5 . Laterality is already observable in very early stages of development: fetuses show coordinated hand movements at 8-12 weeks post-conception with more right than left arm movements in 85% of fetuses [6][7][8] . In children and adults, the prevalence of left-handedness is about 10% 9 . Handedness clearly clusters in families, but its inheritance pattern is not clear and the heritability of handedness is relatively low: approximately 25% in twin studies (with 95% confidence intervals (CI) ranging from 11 to 30% [10][11][12] ) and 11.9% (95% CI 7.2-17.7) based on autosomal Identity by Descent (IBD) information from closely related individuals in the UK Biobank 13 . Early genetic hypotheses on the development of hand preference incorporated a component of randomness 14,15 : depending on which alleles were inherited, a person would be right-handed or have an equal chance of being either left-or right-handed. This randomness has also been referred to as "developmental instability", or "fluctuating asymmetry", representing developmental variance unique to the person 16 . Such randomness could explain monozygotic twin discordance in handedness 15 as reported in some twin studies [17][18][19] , although very early studies did not confirm zygosity by DNA testing.

Results
Epigenome-wide association meta-analysis of left-handedness. Tables 1, 2 and Supplementary   Tables 1-4 display the characteristics of the participants included in the study. The epigenome-wide association study of left-handedness meta-analysed data from adults (N = 3914) with DNA methylation data in peripheral blood (Illumina, 450 k) from NTR (N = 2682, 34% male, mean age at methylation 36.5, standard deviation (SD) 12.7) and ALSPAC (N = 1232, 30% male, mean age at methylation 48.98, SD 5.55). In EWAS discovery cohorts, the prevalence of left-handedness was 12% in NTR, and 8% in ALSPAC. The prevalence of left-handedness as a function of year of birth in NTR is provided in Supplementary Table 2.
We tested 409,562 CpGs with adjustment for age, sex, smoking status, body mass index (BMI), measured or estimated cell proportions, and technical covariates. Genome-wide EWAS test statistics from each cohort separately, and from the meta-analysis, showed no inflation (Supplementary Tables 5-11). None of the associations with CpG sites reached epigenome-wide significance (i.e. Bonferroni adjusted P < 0.05 or false discovery rate < 5%) ( Supplementary Fig. 1).The CpGs with lowest p-values in meta-analysis (P < 1 × 10 -5 ) are shown in Table 3. Six of eight CpGs were located near transcription start sites on different chromosomes: in the LRRC2 gene on chromosome 3, in the ATP6V1B2 gene on chromosome 8, in the CKAP4, GALNTR6, and UNC1198 genes on chromosome 12, in the C13orf18 gene on chromosome 13, in the MBD2 gene on chromosome 18, and in the NTSR1 gene on chromosome 20. The average difference in DNA methylation between left-handers and right-handers at these CpGs was small (from 0.06 to 0.8%; i.e. 0.0006 to 0.008 on the methylation beta-value scale) with lower methylation level in left-handers at all CpGs except for cg13719901 (LRRC2) (Supplementary Figs. 2 and 3).
The DMR meta-analysis detected two DMRs associated with left-handedness (Fig. 2, Table 4, Supplementary Table 12). The DMR on chromosome 20 (BLCAP; NNAT; 16 CpGs) had lower DNA methylation in left-handers than in right-handers (P-value adjusted for multiple testing (P adj ) = 0.00004). The DMR on chromosome 2 (IAH1; 7 CpGs) also had lower DNA methylation in left-handers (P adj = 0.03). In total, 15 of 16 CpGs in the DMR on chromosome 20 (Supplementary Figs. 4 and 5) and 6 of 7 CpGs in the DMR on chromosome 2 (Supplementary Figs. 6 and 7) were hypomethylated in left-handers. The average absolute DNA methylation difference at these regions between left-handers and right-handers based on meta-analysis regression coefficients for the individual CpGs was 0.4% for the DMR on chromosome 20, and 0.1% for the DMR on chromosome 2. Both DMRs were within CpG islands, and were not detected in the individual cohorts. Some CpG sites within lefthandedness-associated DMRs have been previously associated with other traits, these associations are listed in Supplementary Table 13. GWAS follow-up. We tested the overlap of our EWAS meta-analysis results with findings from the most recent GWAS meta-analysis of handedness 13  www.nature.com/scientificreports/ handedness (at P < 5 × 10 -8 ) were on average more strongly associated with left-handedness in the EWAS metaanalysis than the other tested CpGs (β = 0.027, P = 0.04). The effect was weaker when less stringent GWAS p-value cut-offs were applied (i.e. SNPs with P < 1 × 10 -6 , and SNPs with P < 1 × 10 -5 ). Importantly, in a control analysis substituting genetic loci with loci associated with type 2 diabetes 43 , we did not observe a statistically significant overlap (β = 0.005, P = 0.265) (see Supplementary Table 14, Supplementary Fig. 8).
A look-up of left-handedness associated SNPs 13 in the methylation quantitative trait locus (mQTL) database by the Genetics of DNA Methylation Consortium (GoDMC) 44 showed that 95% of left-handedness associated     We repeated the GWAS enrichment analysis with these CpGs driven by mQTLs removed, and obtained similar results, i.e. CpGs near GWAS loci (but not driven by significant mQTLs) were still more strongly associated with left-handedness compared to other genome-wide CpGs (β = 0.027, P = 0.027). None of the CpGs driven by mQTL was located in significant DMRs from our EWAS meta-analysis, or among the top 100 CpGs (by p-value) from our EWAS meta-analysis, which illustrates that our top EWAS findings are not driven by mQTL effects of the top GWAS loci.
Longitudinal analysis. While handedness is a stable trait, DNA methylation can vary over age 45 . We analyzed DNA methylation in ALSPAC offspring measured in cord blood at birth, and in peripheral blood at 7, 17, and 24 years old (N = 791, Table 2 and Supplementary Table 3 Table 4, Supplementary  Fig. 13). These DMRs did not overlap with DMRs detected in the analyses of blood methylation data. Sixteen of 18 CpGs from these regions had a lower methylation level in left-handed children than in right-handed (Supplementary Table 29).

Sensitivity analyses.
We reported the DNA methylation and left-handedness association study with adjustment for prenatal and postnatal factors that influence DNA methylation as shown in previous studies: BMI 38 and smoking 46 in adults, and gestational age, birth weight and prenatal maternal smoking in children 47,48 . We examined if the EWAS results for handedness differ without taking these factors into account. Across all analyses, the correlations between the effects for the top 100 CpGs were strong between the models with and without adjustment for these factors (r ranged from 0.99 to 1), and overlaps of the top 100 CpGs were substantial . Adjustment for the factors increased the number of DMRs associated with left-handedness in meta-analysis (1 DMR without adjustment and 2 DMRs with adjustment), and in EWAS in children (2 DMRs without adjustment, and 4 with adjustment in buccal cells in NTR).  Tables 15-18). We compared the methylation results obtained in discordant MZ twins to the EWAS meta-analysis results for the top 100 CpGs ranked on ascending P-value from each analysis. To avoid sample overlap, we repeated the EWAS meta-analysis after exclusion of the MZ discordant twin pairs. Correlations between the meta-analysis top 100 effects and mean methylation differences from the within-pair analysis were weak in adults (r MZ disc adults blood-Meta-analysis = 0.189, P = 0.06; r Meta-analysis-MZ disc adults blood = 0.188, P = 0.06, α = 0.0003) and children (r MZ disc children buccal-Meta-analysis = 0.134, P = 0.18; r Meta-analysis-MZ disc children buccal = 0.252, P = 0.01) ( Supplementary Fig. 10). There were few overlapping CpGs among the top 100 CpGs from the within-pair analyses and other analyses ( Supplementary Fig. 11).

Handedness methylation scores.
To examine if variation in handedness can be predicted by DNA methylation levels across multiple CpGs, methylation scores (MS) were created. These were based on EWAS summary statistics in NTR to predict into ALSPAC, and on ALSPAC summary statistics to predict into NTR given the following p-value thresholds to include CpGs: P < 1 × 10 -1 , P < 1 × 10 -3 , P < 1 × 10 -5 . To estimate the variance explained by MS above genetic variants, polygenic scores (PGS) were created based on the summary statistics from the handedness GWAS of Cuellar-Partida et al. 13 with exclusion of NTR, ALSPAC and 23andMe cohorts (N GWAS = 196,419). Since four scores were tested (3 methylation scores, one polygenic score), we applied Bonferroni correction for four tests (α = 0.05/4 = 0.0125) (Supplementary Table 30, Supplementary Fig. 14). The results are summarized in Supplementary Table 31. MS did not predict left-handedness in NTR and ALSPAC adults, or in children at 7-9 years old and did not explain variance over and above the variance explained by the PGS in the combined model (R 2 MS from -0.17 to 1.28%, R 2 PGS from 0.002 to 0.46%). The largest amount of explained variance was in ALSPAC at 7 years old for the MS based on CpGs at P < 1 × 10 -5 (R 2 MS = 1.28%, P = 0.1, N CpGs = 7).

Discussion
We have performed an epigenome-wide association study of left-handedness, including left-and right-handed individuals from two population-based cohorts from the Netherlands and the UK. In the meta-analysis, combining the NTR and ALSPAC adult cohorts, two DMRs associated with left-handedness. The first DMR (genomic location: chr20q11.23, 36,148,679:36,149,022) is located within the 5'UTR of the BLCAP apoptosis inducing factor (BLCAP) gene and nearby the transcription start site (TSS1500) of the neuronatin (NNAT) gene. BLCAP encodes a protein that reduces cell growth by stimulating apoptosis. NNAT is located within intron of BLCAP and is involved in brain development and nervous system structure maturation and maintenance. BLCAP and NNAT are imprinted in the brain 49 . The second intron of NNAT regulates the expression of BLCAP transcripts acting as an imprint control region regulating also allele-specific DNA methylation and histone modifications 50 . Around 50% of CpGs in a region covering the NNAT CpG islands are methylated in brain and other tissues, suggestive of differential allele-specific methylation 51 . The imprinted DMR for these genes [chr20: 36,139,941-36,159,190 52 ] overlaps with the left-handedness DMR that we identified (chr20:36,148,679-36,149,022). A potential connection of genomic imprinting with handedness was previously suggested in a study of another imprinted gene LRRTM1 41 . CpGs from the handedness-associated region at chromosome 20 previously associated with myalgic encephalomyelitis and chronic fatigue syndrome, preterm birth, obesity, metabolic parameters, and arm fat mass (DXA scan measurement). The second DMR (genomic location: chr2p25.1, 9,614,471:9,614,744) is located within the isoamyl acetate hydrolyzing esterase 1 (IAH1) gene. The IAH1 gene encodes an acyl esterase and is associated with neonatal inflammatory skin and bowel disease, and a disease with an inborn error of leucine metabolism (3 methylglutaconic aciduria type 1). CpGs from the region previously associated with gestational age, bone mineral density, metabolic parameters, and schizophrenia. Some of these traits have been reported to be associated with handedness in epidemiological studies, e.g. BMI 53 and gestational age 54 , for which we adjusted in our analyses. Previous analysis of the genetic correlations between left-handedness and 1349 complex traits using LD-score regression did not reveal any genetic correlations at FDR < 5%, but suggestive positive correlations were observed with neurological and psychiatric traits, including schizophrenia 13 .
Even though no DMPs were identified after correction for multiple testing, and effect sizes of top CpGs were small (mean differences between left-and right-handed individuals smaller than 1%), the high-ranking CpGs are of potential interest. The second-ranking CpG cg09239756 (genomic location: chr12, 106,642,360) is located near the cytoskeleton associated protein 4 (CKAP4) gene. This gene mediates the anchoring of the endoplasmic reticulum to microtubules. Microtubules are an important cytoskeleton component that play a role in neuronal morphogenesis and migration, and axon transport 31 . Microtubules have been widely discussed in association with handedness 15,29 and brain anatomical asymmetry 32 , and genes involved in microtubule pathways were enriched in the GWAS of handedness 13 . Moreover, in our enrichment analysis, we found that CpGs located within a 1 Mb window from SNPs associated with left-handedness in the GWAS meta-analysis by Cuellar-Partida et al. 13 were more strongly associated with left-handedness in our meta-analysis compared to CpGs outside of this window. Larger EWAS meta-analysis or replication in additional independent cohorts is necessary to establish the robustness of the top DMPs. www.nature.com/scientificreports/ Hand movements together with other lateralized movements and molecular signs of lateralization are observed at very early stages of human development in the uterus [6][7][8] . Therefore, DNA methylation differences associated with hand preference are expected to emerge early in development. While DNA methylation at some CpGs in the genome changes throughout the lifespan 45 , the DNA methylation signal associated with left-handedness was moderately consistent from birth throughout the lifespan: DMP effects correlated in ALSPAC offspring from birth to 24 years old, although genome-wide significance for DMPs was not reached. Consistency in DNA methylation signal associated with left-handedness at different time-points may indicate that the pattern for left-handedness is conserved through the lifespan.
Several DMRs were detected in buccal cells in children around 9 years old (genomic locations: chromosomes 8, 22, 9, and 12) after correction for multiple testing. Annotation of these regions implicate the following protein coding genes: the plectin gene (PLEC1), the myoglobin gene (MB) gene, the elongator complex protein 1 gene (ELP1), the actin binding transcription modulator gene (ABITRAM), and the WNK lysine deficient protein kinase 1 gene (WNK1). The CpGs in these regions are mostly hypomethylated in left-handed individuals. The genes encode for proteins that participate in cytoskeleton functions, chromatin organization, development of neurons, and metabolism. CpGs from DMRs in buccal cells previously associated with other phenotypes: CpGs from the DMR on chromosomes 8 with myalgic encephalomyelitis, chronic fatigue syndrome, multiple sclerosis, and gestational age; CpGs from the DMR on chromosome 9 with bone mineral density, tissue mass of the arm (DXA scans measurement), and multiple metabolic parameters. Interestingly, some of these phenotypes also associated with CpGs from our meta-analysis in blood DNA methylation.
A difference in handedness preference in MZ twin pairs has always fascinated parents of twins and twins themselves: how can children with almost identical genes differ for such a prominent trait? Handedness discordance in identical twins was described a long time ago 17,18 , and the percentage of MZ discordant twins were reported as 20% of 3486 MZ twins in East Flanders 19 , and 19% of 1724 MZ twins from a London twin study 28 . We observed that 21% of adult MZ twins and 24% of young MZ twins were discordant for handedness in our study, but we did not detect DNA methylation differences among them in blood or buccal cells. This null finding could mean that handedness discordance is not associated with methylation differences in the tissues that we studied (but might be present in other tissues), or that our analysis was underpowered to detect methylation differences associated with handedness discordance. Our discordant MZ twin analysis may be underpowered to detect small DNA methylation differences 55 , as it included only 133 MZ discordant adult twin pairs and 86 child twin pairs. Different, not mutually exclusive, hypotheses have been proposed for handedness discordance of MZ twins, including large unique environmental effects, that are deduced based on the low heritability of around 25% as estimated in the meta-analysis by Medland et al. 10 . McManus 16 emphasizes that such unique environmental factors may represent what in biology is referred to as noise, randomness or fluctuating asymmetry. McManus quotes Mitchell 56 saying that such processes are "caused not by any factors outside the organism, but by inherent variation in the processes of development". These reflect developmental variance unique to the person. In a discussion and review of the human and animal studies literature Molenaar et al. 57 called these processes 'the third source of individual differences' , i.e. a third source besides genetic and environmental influences on individual differences and discuss how deterministic growth process may give rise to highly variable results.
There is a growing interest to improve the prediction of traits with use of other omics data than SNPs, like DNA methylation 37 and by non-genetic early life factors (e.g. earlier factors associated with left-handedness including birth weight, being multiple, month of birth, breastfeeding etc. 58 ). Given the low heritability of handedness (~ 25% 10,11 ), it is expected that non-genetic factors play a role. Although together early life factors had minimal predictive value 58 , they may inspire the search for DNA methylation signatures, as DNA methylation signatures later in life were found for birthweight 47 . Single CpGs did not individually reach statistical significance in our EWAS, but combining information across multiple CpGs into an overall methylation score can be a more powerful approach to capture variation in handedness. We calculated methylation scores as weighted sums of the individual's methylation loci beta values of a pre-selected number of CpG sites. However, the predictive value of polygenic and methylation scores for handedness was low, which likely reflects that current GWAS and EWAS analyses for handedness are still underpowered.
Our multi-cohort epigenome-wide association study can be summarized in several key steps presented in Fig. 3. We examined DNA methylation data in different tissues (whole blood, cord blood, buccal cells) and ages (from birth to adulthood). The limitations of the study are related to handedness measurements, available tissues, differences in platforms used for DNA methylation (Illumina 450 k, EPIC), and study power. There were slight differences in the assessment of left-handedness between NTR and ALSPAC. Power might increase by investigating a refinement of the handedness phenotype (e.g. hand skill measurement rather than self-report of the preferred hand), analysis of DNA methylation in more relevant tissues, and an increase in sample size. The difference in left-handedness rates among children born before and after 1960 may be due to a move away from being forced to use the right hand prior to 1960 59 . We accounted for this trend by including age (which correlates almost perfectly with birth year in these samples) as a covariate in the analyses, however, it should be noted that the forced use of the right hand in older generations may render the phenotype definition of handedness less precise. Our meta-analysis was based on whole blood methylation data, where the methylation level represents the overall level of DNA obtained from millions of white blood cells. We observed methylation differences between left-and right-handed individuals of up to 0.8% at top-DMPs. This small effect size could reflect that a methylation difference is present in only a sub-set of cells in an individual, or a sub-set of individuals in the population, or a combination. The biological implications of these findings remain to be established and our top-DMPs remain to be replicated in additional cohorts or larger meta-analysis. The primary tissues of interest for handedness are brain 2,29 , spinal cord 40 , and arm muscle tissues 4 , and the timing when these tissues are collected could also play a role, but the collections of these tissues are not widely available in population cohorts for obvious reasons. Although 4000 is considered a decent sample size for EWAS and smaller sample sizes have  Figure 3. Seven-step approach to DNA methylation signature discovery, incorporating twin design. The figure represents the methodology of DNA methylation signature discovery study of a phenotype in multiple steps. It integrates behavioral-genetic and SNP-based methods (step 1) to estimate heritability, epigenome-wide study methods (steps 3-4) for association analyses, follow-up of results using summary statistics from previous EWASs and GWASs (step 5), the discordant twin design (step 6), and methods integrating polygenic and DNA methylation data (step 7) in enrichment analysis. Specific methods for each step are presented on the left and outcomes on the right. GWAS, genome-wide association study. EWAS, epigenome-wide association study. SNP, single nucleotide polymorphisms. www.nature.com/scientificreports/ allowed for the successful detection of many loci where DNA methylation robustly associates with traits such as body mass index 38 , the effect sizes for handedness were unknown a priori and 4000 may still be too small to detect methylation differences associated with left-handedness. Similarly, sample sizes of EWASs of behavioral and psychiatric traits are now increasing beyond 10,000 60 . For GWAS of handedness that applied similar phenotyping as the current study, the increase of the sample size from about 400,000 to 1.7 million increased the number of associated loci from a handful to over forty 13 , and similar increases in the number of detected loci may occur when EWAS sample sizes for left-handedness increase. Finally, our study focused on DNA methylation, but other epigenetic processes could play a role in handedness such as histone modifications, post-translational regulation by miRNAs and X-chromosome inactivation that remain to be explored. We reported an EWAS of left-handedness in two large population-based cohorts with data from children and adults, and examined performance of methylation scores and polygenic scores. Despite the plausible rationale of multiple genetic and non-genetic factors that may act via epigenetic pathways to influence the development of handedness, we did not uncover support for the hypothesis that DNA methylation in peripheral tissues captures much if any of the variation in handedness. We propose that future studies consider other tissues, such as related to central nervous system.

Methods
Overview. The primary epigenome-wide association study (EWAS) of left-handedness was performed in two cohorts with DNA methylation data in whole blood (Illumina, 450 k): NTR adults 61 (N = 2682 individuals including twins, mean age at methylation 36.5, SD 12.7), and ALSPAC adults 62,63 (N = 1232, mean age at methylation 48.98, SD 5.55). EWAS analyses were performed in each dataset separately, and summary statistics were combined in the meta-analysis (N = 3914) testing 409,563 CpGs. As this is a meta-analysis of existing DNA methylation datasets, no analyses were done to pre-determine sample sizes, but the sample size is larger compared to previously published DNA methylation studies of handedness 41,42 . We tested whether the EWAS signal was enriched in nearby loci detected in the previous GWAS on handedness 13 . Secondary analyses were performed in different tissues: in cord blood and peripheral blood in ALSPAC offspring 64 , i.e. the children of ALSPAC participants that contributed to the primary EWAS (N = 791 with DNA methylation data at birth, at 7, and 17 years old (Illumina 450 k chip), and/or at 24 years old (Illumina EPIC array)), and in buccal cells from an independent group of children from the NTR 65,66 (N = 946 twins, mean age 9.5, SD 1.85, Illumina EPIC array). The number with DNA methylation and covariate data in ALSPAC differed at different time points from 442 to 759. We carried out within-pair twin analysis in NTR MZ twins discordant for handedness (N adults = 133 twin pairs, N children = 86 twin pairs). We performed EWAS analyses in each dataset and examined correlations between the regression coefficients of top CpGs (ranked on ascending p-value) from each EWAS analysis. Finally, we created and tested polygenic and DNA methylation scores for left-handedness. The study method and design are presented in Figs. 1 and 3. Detailed cohort information is provided in Appendix 1.
Handedness. NTR. Information on hand preference for adults and children was collected by surveys and in small subgroups from laboratory-based projects. Parental reports on children were collected at 5 years and included seven items for different activities, from which the item "What hand does child use for drawing?" was selected. The four answer categories were left-handed, right-handed, both hands and do not know. Multiple adult surveys included the question: "Are you right-handed or left-handed?" (4 surveys) or "Are you predominantly left-handed or right-handed?" (3 surveys). The three answer categories were left-handed, right-handed, and both. For a small number of adult self-reports at younger ages (14, 16 or 18 years) or parental assessment at age 5 were also available.
ALSPAC. Adults (mothers and fathers) were asked which hand they used to write, draw, throw, hold a racket or bat, brush their teeth, cut with a knife, hammer a nail, strike a match, rub out a mark, deal from a pack of cards or thread a needle (11 questions). Child handedness was assessed at 42 months by questionnaire in which the mother was asked which hand the child used to draw, throw a ball, color, hold a toothbrush, cut with a knife, and hit things (6 questions). Responses were scored − 1, 0 or 1 for left, either or right, respectively. Handedness was coded as 1 for left-handed or 0 for right-handed in both cohorts.

DNA methylation and genotyping. DNA methylation was measured with the Infinium HumanMeth-
ylation450 BeadChip Kit which measures more than 450,000 methylation sites (primary analysis in adults in NTR and ALSPAC and secondary analysis in ALSPAC offspring at birth, 7 and 17 years old), or the Infinium MethylationEPIC BeadChip which measures more than 850,000 methylation sites (secondary analysis in ALSPAC offspring at 24 years old and NTR children). Genotyping for polygenic risk scores was done on multiple platforms with imputation of the target data using reference haplotypes from 1000 genomes reference panel. Cohort-specific details on biosample collection, DNA methylation profiling, quality control, cell-type proportions measurements and genotyping are described in Appendix 1.
Intergroup differences. We tested if there were differences in characteristics that were included in EWAS models (such as age at biological sample collection, sex, body mass index (BMI), smoking status at blood collection for adults, and gestational age, maternal smoking during pregnancy, birth weight for children, cell proportions/percentages in buccal swabs and in blood samples) between left-and right-handed individuals by generalized estimating equations (GEE) to accommodate the relatedness among the twins in NTR, and by standard logistic regression in ALSPAC. The R package 'gee' was used with the following specifications: binomial (for ordinal data) link function, 100 iterations, and the 'exchangeable' option to account for the correlation structure www.nature.com/scientificreports/ within families and within persons. Right-and lefthanded MZ discordant twins were compared with paired t-test for the traits that were not identical in twins (birth weight, BMI, smoking, cell percentages). All statistical tests here and below were two-tailed.

Epigenome-wide association analyses. Primary analyses.
The association between DNA methylation levels and left-handedness was tested for each site under a linear model (ALSPAC) or generalized estimating equation (GEE) model accounting for relatedness of twins (NTR). DNA methylation beta-values were the dependent variable and were typically normally distributed. The following predictors were included in the basic model: handedness (coded as 0 = right-handed and 1 = left-handed), sex, age at blood sampling, percentage of blood cells for blood samples, and technical covariates in NTR and ALSPAC (see Appendix 1). An adjusted model was fitted to account for BMI and smoking status at blood draw in both NTR and ALSPAC adult cohorts, because BMI and smoking have large effects on DNA methylation in adults 38,46 . The primary results reported in the paper are based on the fully adjusted model. The models are described in Appendix 2. Throughout the text, we refer to regression coefficients from the EWAS, which represent the methylation difference between lefthanded and right-handed individuals on the methylation beta-value scale. A positive regression coefficient (β) means a higher methylation level in left-handed individuals. The value of an individual on the methylation scale is commonly also symbolized as beta (β) and ranges from 0 to 1, where 0 represents a methylation level of 0% and 1 represents a methylation level of 100%.
Secondary analyses. The same basic models were fitted to the data from ALSPAC and NTR children. For DNA methylation in buccal cells, cell proportions (epithelial cells, natural killer cells) for buccal samples were included instead of percentage of blood cells. As several characteristics, such as gestational age and birthweight, affect DNA methylation 48,67 , we included these in the adjusted model in children (see Appendix 2).
In the within-pair analysis of discordant MZ twins, paired t-tests were employed to test for methylation differences between the left-handed and the right-handed twins. Paired t-tests were performed in R on residual methylation levels, which were obtained by adjusting the DNA methylation β-values for sample plate, array row, cell proportions in buccal samples in children and sample plate, array row, and percentages of blood cells in adults. Additional covariates, birth weight in children and BMI and smoking status in adults, were added in adjusted model. Age, sex, maternal smoking, and gestational age were not included because these variables are identical in MZ twins.
To account for multiple testing, we considered Bonferroni correction and a False Discovery Rate (FDR) of 5%. The Bonferroni corrected p-value threshold was calculated by dividing 0.05 by the number of genome-wide CpGs tested, and false discovery rate (FDR) q-values were computed with the R package 'qvalue' with default settings. The Bayesian inflation factor (λ) was calculated with the R package Bacon 68 (see Supplementary Table 5).

Meta-analysis.
A meta-analysis was performed in METAL 69 based on estimates (regression coefficients) and standard errors from the EWAS of handedness performed with GEE in NTR and linear regression in ALSPAC. NTR and ALSPAC adult cohorts were combined. In total, 409,563 CpG sites present in both cohorts were tested with statistical significance evaluated after Bonferroni correction and at an FDR q-value < 0.05.

Comparison of top CpGs from different analyses. To compare top CpGs from different analyses,
we repeated the NTR EWAS analyses in adults in children and meta-analysis with discordant MZ twin pairs removed to avoid sample overlap. We selected methylation sites that overlapped in 13 analyses with adjusted model (meta-analysis, meta-analysis without discordant MZ twins, EWAS NTR adults, EWAS NTR adults without discordant MZ twins, EWAS ALSPAC adults, EWASs ALSPAC at birth, 7, 17, 24 years, EWAS NTR children, EWAS NTR children without discordant twins, and within-pair analyses of discordant MZ twin adults and children) that resulted in 379,924 methylation sites. We calculated Pearson correlations for effect estimates of the top 100 CpGs ranked by p-value from one analysis with the effect estimates of the same CpGs in other analyses. Statistical significance of correlations was assessed after Bonferroni correction for the number of correlations tested: α = 0.05/(13 × 13 − 13) = 0.0003.

Differentially methylated regions.
We used the R dmrff library 36 for R to identify regions where CpG sites showed evidence for association with handedness. Dmrff identifies DMRs by meta-analysing EWAS summary statistics from CpG sites in each region while adjusting for dependencies between the sites and uncertainty in the EWAS effects (https:// github. com/ peris hky/ dmrff). In this study, dmrff was applied separately in the NTR and ALSPAC cohorts, and then used to identify DMRs in common between the cohorts by meta-analysis. As previously described 36 , DMR meta-analysis preceded by first identifying candidate DMRs using the EWAS metaanalysis summary statistics, calculating DMR statistics for these candidates in each cohort separately, and then meta-analysing the DMR statistics across the two cohorts. The DMR effect size is a weighted sum of the EWAS effects for each CpG site (i.e. methylation differences between LH and RH). All dmrff p-values were adjusted for multiple tests (Bonferroni adjustment) by multiplying them by the total number of DMRs considered. We report significant regions (P adj < 0.05) including at least two CpG sites within a 500 bp window observed to be nominally associated with handedness by EWAS (P < 0.05). The average absolute DNA methylation difference in the region between left-handers and right-handers is calculated as the sum of absolute regression coefficients of each CpG in the region divided by the number of CpGs. We plotted the DMRs with the coMET R Bioconductor package 70 to graphically display additional information on physical location of CpGs, correlation between sites, statistical significance, and functional annotation (annotation tracks included genes Ensembl, CpG islands (UCSC), regulation Ensembl). We looked up CpG sites associated with handedness-associated SNPs with a P-value < 1.0 × 10 -08 (based on the GWAS meta-analysis by Cuellar-Partida et al. 13 without 23andMe, NTR and ALSPAC, resulting in a list of 420 SNPs) using the mQTL database maintained by the Genetics of DNA Methylation Consortium (GoDMC, N = 27,750 European samples; http:// mqtldb. godmc. org. uk/ about) 44 . We then checked if associations with handedness were observed for these sites in our EWAS meta-analysis and DMR meta-analysis.
Polygenic and methylation scores. Polygenic scores (PGS) for handedness were calculated based on the GWAS meta-analysis without 23andMe by Cuellar-Partida et al. 13 . To avoid overlap between the discovery and target samples, summary statistics without NTR and ALSPAC were requested (196,419 individuals, N SNPs = 13,550,404). The linkage disequilibrium (LD) weighted betas were calculated using a LD pruning window of 250 KB, with the fraction of causal SNPs set at 0.50 by LDpred 73 . We randomly selected 2500 2nd degree unrelated individuals from each cohort as a reference population to calculate the LD patterns. The resulting betas were used to calculate the PGSs in each dataset using the PLINK 1.9 software. All PGSs were standardized (mean of 0 and standard deviation of 1). Methylation scores (MS) were calculated in NTR based on EWAS summary statistics obtained from ALSPAC, and vice versa, as previously done to create methylation scores for BMI and height 74 . We calculated same-tissue same-age DNA-methylation scores based on methylation data from NTR adults (blood) and ALSPAC parents (blood), and cross-tissue DNA-methylation scores based on data from NTR and ALSPAC offspring, with DNA methylation measured in buccal cells, and blood, respectively (see Fig. 1). For each individual, a weighted score sum was calculated for left-handedness by multiplying the methylation value at a given CpG by the effect size of the CpG (β), and then summing these values over all CpGs: DNA methylation score (i) = β 1 *CpG 1i + β 2 *CpG 2i ··· + β n *CpG ni , where CpG n is the methylation level at CpG site n in participant i, and β n is the regression coefficient at CpG n taken from summary statistics of the EWAS analysis. All methylation scores were standardized (mean of 0 and standard deviation of 1). We used weights from summary statistics of EWASs in four cohorts: NTR adults, ALSPAC adults, NTR children, ALSPAC offspring at 7 years old. Subsets of CpGs to be included in methylation scores were selected based on P-value < 1 × 10 -1 , < 1 × 10 -3 , and < 1 × 10 -5 . We analysed the predictive value of the left-handedness polygenic scores and methylation scores in NTR and ALSPAC adult and child cohorts from our EWAS study. To quantify the variance explained by the PGS and MS, we used the approach proposed by Lee et al. 75 , where coefficients of determination (R 2 ) for binary responses are calculated on the liability scale. The equations of all models are provided in Appendix 2. Statistical significance was assessed following Bonferroni correction for the number of scores tested (PGS and 3 MSs). This resulted in α = 0.05/4 = 0.0125, nominal significance at 0.05.
Ethics statement. All methods were performed in accordance with the Declaration of Helsinki. For NTR, the study was approved by the Central Ethics Committee on Research Involving Human Subjects of the VU University Medical Centre, Amsterdam, an Institutional Review Board certified by the U.S. Office of Human Research Protections (IRB Number IRB00002991 under Federal-wide Assurance FWA00017598; IRB/institute codes, NTR 03-180). All subjects provided written informed consent. For children, written informed consent was given by their parents. For ALSPAC, ethical approval for the study was obtained from the ALSPAC Ethics and Law Committee and the Local Research Ethics Committees. All subjects provided written informed consent. For children, written informed consent was given by their mothers.