Data from: BsRADseq: screening DNA methylation in natural populations of non-model species

Epigenetic modifications are expected to occur at a much faster rate than genetic mutations, potentially causing isolated populations to stochastically drift apart, or if they are subjected to different selective regimes, to directionally diverge. A high level of genome-wide epigenetic divergence between individuals occupying distinct habitats is therefore predicted. Here, we introduce bisulfite-converted restriction site associated DNA sequencing (bsRADseq), an approach to quantify the level of DNA methylation differentiation across multiple individuals. This reduced representation method is flexible in the extent of DNA sequence interrogated. We showcase its applicability in three natural systems, each comprising individuals adapted to divergent environments: a diploid plant (Heliosperma, Caryophyllaceae), a tetraploid plant (Dactylorhiza, Orchidaceae) and an animal (Gasterosteusaculeatus, Gasterosteidae). We present a robust bioinformatic pipeline, combining tools for RAD locus assembly, SNP calling, bisulfite-converted read mapping and DNA methylation calling to analyse bsRADseq data with or without a reference genome. Importantly, our approach accurately distinguishes between SNPs and methylation polymorphism (SMPs). Although DNA methylation frequency between different positions of a genome varies widely, we find a surprisingly high consistency in the methylation profile between individuals thriving in divergent ecological conditions, particularly in Heliosperma. This constitutive stability points to significant molecular or developmental constraints acting on DNA methylation variation. Altogether, by combining the flexibility of RADseq with the accuracy of bisulfite sequencing in quantifying DNA methylation, the bsRADseq methodology and our bioinformatic pipeline open up the opportunity for genome-wide epigenetic investigations of evolutionary and ecological relevance in non-model species, independent of their genomic features.

表观遗传修饰（epigenetic modifications）的发生速率预计远高于基因突变（genetic mutations），其可导致隔离种群发生随机趋异；若种群面临不同的选择机制，则会发生定向分化。据此我们预测，栖息于不同生境的个体之间，全基因组范围内的表观遗传分化水平将处于较高水平。本研究引入亚硫酸氢盐转化限制性酶切位点关联DNA测序（bisulfite-converted restriction site associated DNA sequencing，bsRADseq）方法，用于量化多个个体间的DNA甲基化分化程度。该简化代表性测序方法可灵活调整所检测的DNA序列范围。我们在三类自然系统中验证了该方法的适用性，每类系统均包含适应不同环境的个体：二倍体植物（Heliosperma，石竹科（Caryophyllaceae））、四倍体植物（Dactylorhiza，兰科（Orchidaceae））以及动物（Gasterosteusaculeatus，刺鱼科（Gasterosteidae））。我们构建了一套稳健的生物信息学分析流程，整合了RAD位点组装、单核苷酸多态性（SNP）检测、亚硫酸氢盐转化读段比对与DNA甲基化检测工具，可用于分析有无参考基因组的bsRADseq数据。尤为关键的是，本方法可准确区分单核苷酸多态性与甲基化多态性（SMPs）。尽管基因组不同位点的DNA甲基化频率差异显著，但我们发现，栖息于不同生态环境的个体之间，其甲基化谱却具有惊人的高度一致性，在Heliosperma类群中尤为突出。这种组成型稳定性表明，DNA甲基化变异受到了显著的分子或发育约束。综上，通过将RADseq的灵活性与亚硫酸氢盐测序在DNA甲基化定量上的准确性相结合，本研究提出的bsRADseq方法与生物信息学流程，为非模式物种开展与进化和生态相关的全基因组表观遗传研究提供了可行路径，且不受其基因组特征的限制。