About 7% of Epigenetic Inheritance Patterns in Mice Were Non-Mendelian; Paramutation Confirmed at Capn11
Background
An analysis of DNA methylation inheritance across generations in mouse liver and muscle found that about 7% of the autosomal epigenetic inheritance patterns identified in the study did not follow Mendelian laws. Davidovich and colleagues used long-read nanopore sequencing, which reads long strands of DNA, to investigate epigenetic inheritance across the genome. The study was published in Nature Genetics.
Epigenetic phenomena, including genomic imprinting, show that inheritance at the molecular level can depart from typical Mendelian patterns. Yet how DNA methylation is transmitted from parents to offspring remains incompletely understood. The study asked to what extent methylation inheritance follows Mendelian laws and what patterns fall outside them.
Key findings
About 93% of the autosomal epigenetic inheritance patterns identified in the study followed Mendelian laws. The researchers explained that these patterns were governed mainly by cis-acting methylation quantitative trait loci. In other words, genetic factors acting near the regions where methylation occurs largely explain the inheritance patterns.
About 7% showed non-Mendelian inheritance patterns. This proportion was calculated among the autosomal epigenetic inheritance patterns identified in the study. It does not refer to 7% of all bases in the genome or all methylation sites. The abstract also did not conclude that the environment or parental state caused these inheritance patterns.
The researchers reported newly arising epigenetic inheritance patterns and widespread sex-specific DNA methylation patterns restricted to the liver. They also proposed five putatively novel imprinted genes on the autosomes and X chromosome. Because the researchers described their novelty as tentative, this should not be interpreted as the definitive discovery of multiple new imprinted genes.
In particular, naturally occurring intergenerational paramutation was confirmed at a lineage-specific transposable element within Capn11. The same phenomenon was reported as likely at Vps37c. The evidence levels should be distinguished: Capn11 is a confirmed case, whereas Vps37c is a likely case. The abstract alone does not provide enough information to explain the detailed molecular mechanism behind this phenomenon or its effects on gene expression.
Meaning and outlook
These findings provide a basis for considering patterns of epigenetic inheritance alongside DNA sequence when analyzing inherited traits and disease. Most patterns followed Mendelian laws, but some showed other modes of inheritance. The sex-specific methylation reported in the liver also points to the need for interpretations that distinguish tissue and sex.
However, the materials reviewed here were a bibliographic record and abstract. The number of animals, breeding design, observation period, results for individual genes, and detailed validation procedures could not be confirmed. These materials alone also do not establish how broadly the Capn11 observation can be generalized to other loci or tissues.
Improved prediction of human disease risk or diagnostic performance was not validated in these materials. Applying the approximately 7% observed in mice as a correction factor in a human disease prediction model would require separate evidence. Follow-up studies should examine the reproducibility of the observed inheritance patterns and their underlying mechanisms in detail.
The account of the study is based on the provided Europe PMC bibliographic record and abstract. The paper is available at the Nature Genetics DOI.
Nature Genetics, Published online: 20 May 2026. DOI: 10.1038/s41588-026-02604-z
Summary: Breaking the traditional dogmas of inheritance, Davidovich et al. utilize high-resolution allele-specific DNA methylation sequencing to map epigenetic transmission landscapes in mouse liver and muscle tissues. While the baseline architectures adhere to Mendelian segregation, a distinct ~7% of the methylome operates via non-Mendelian dynamics. The framework uncovers a novel inventory of imprinted genes and captures a definitive paramutation at the Capn11 locus, where non-coding epigenetic silencing is transferred across homologous alleles. This metadata establishes a programmable baseline for adjusting multi-omic disease susceptibility scoring models.
Researchers studying the inheritance of methylation across generations can use these findings to design analyses that distinguish patterns associated with genetic factors, sex- and tissue-specific patterns, and non-Mendelian patterns. For example, comparing liver and muscle results while also stratifying by sex may help avoid interpreting distinct inheritance patterns as a single average. The observation at the transposable element within Capn11 provides a specific candidate case for further investigation of intergenerational epigenetic inheritance. The full paper and methods should be reviewed before designing experiments. Evidence to support applying these findings to human disease-risk calculations or improving the performance of commercial prediction models has not yet been established.