DNA methylation is involved in regulation of gene expression and hence plays a crucial role in cell differentiation and development. When a gene encoding a protein which is part of the epigenetic machinery (for example proteins that add or remove epigenetic marks, and thereby controls the expression of other genes) is mutated, this can result in alterations in the genome-wide DNA methylation pattern, affecting expression of multiple genes. These patterns (so-called DNAm signatures) were found to be disease-specific, especially for the disorders of epigenetic machinery. DNAm signatures can therefore be utilized to determine the pathogenicity of gene variants through training an artificial intelligence (AI) model to recognize this specific DNAm signature. The resulting AI-model is then applied to determine whether the DNAm pattern of a person with a given gene variant matches the DNAm signature specific to the disorder this gene is associated with. With this it is possible to evaluate whether a given variant is pathogenic or not. DNAm signatures have so far been discovered for over 100 different rare genetic disorders and have become a highly valuable tool in genetic diagnosis. We will present our work with DNAm signatures.