Back to research

Research paper

Episignatures in practice: independent evaluation of published episignatures for the molecular diagnostics of ten neurodevelopmental disorders

Independent validation of published episignatures. Specificity 100% across all signatures. DNMT3A, ATRX, KMT2D, and NSD1 episignatures reach 100% sensitivity. CREBBP-RSTS and one CHD8 signature reach <40% sensitivity — an important caveat for interpreting negative results.

Indexed context

van der Sanden BPGH, et al.

episignaturemethylationdnmt3ansd1diagnostic-sensitivity

Markdown path

content/research/papers/2023-van-der-sanden-episignature-sensitivity.md

Findings

Independent validation of published episignatures. Specificity 100% across all signatures. DNMT3A, ATRX, KMT2D, and NSD1 episignatures reach 100% sensitivity. CREBBP-RSTS and one CHD8 signature reach <40% sensitivity — an important caveat for interpreting negative results.

Why it may matter for Levi

The single strongest external argument for prioritizing an episignature panel in Levi's workup: DNMT3A/TBRS and NSD1/Sotos episignatures both have essentially perfect clinical sensitivity in independent hands. A negative blood episignature at DNMT3A would be a strong functional argument against TBRS that the existing WGS cannot provide.

Paper text

Episignatures in practice: independent evaluation of published episignatures for the molecular diagnostics of ten neurodevelopmental disorders

van der Sanden BPGH, et al. — European Journal of Human Genetics (2023). https://www.nature.com/articles/s41431-023-01474-x

Findings summary

Independent head-to-head validation study of published peripheral-blood DNA methylation episignatures across ten neurodevelopmental disorders, using 101 carriers of (likely) pathogenic variants, 57 VUS carriers, and 25 healthy controls. Central operational conclusions:

  • Specificity across all tested episignatures was 100% (no false positives among healthy controls).
  • Sensitivity varied enormously by disease. DNMT3A (TBRS), ATRX, KMT2D (Kabuki 1), and NSD1 (Sotos) signatures all reached 100% sensitivity in this independent validation.
  • Several other published signatures performed far worse in independent hands (CREBBP / Rubinstein-Taybi <40%, one CHD8 signature <40%), which is an important caveat for how to interpret a negative episignature result for any given gene.
  • The authors conclude episignature testing is clinically useful and should be considered as a testing-cascade step, particularly for variant reclassification and for unexplained suspected chromatinopathy phenotypes.

Relevance to Levi

This paper matters for one specific reason in Levi's case: the DNMT3A / TBRS episignature specifically has essentially perfect clinical sensitivity in independent hands. That means:

  • If Levi has a clinically significant disruption of DNMT3A function — whether from a coding variant missed by short-read sequencing, a deep-intronic or regulatory variant, a structural variant at the locus, or another mechanism that produces the same downstream methylation outcome — a blood episignature assay is expected to detect it.
  • A negative DNMT3A episignature in blood would therefore be a strong argument against the TBRS hypothesis in a way that the existing negative WGS cannot provide, because WGS with standard germline interpretation does not functionally test DNMT3A activity.
  • The same platform simultaneously tests NSD1 (Sotos, 100% sensitivity in this study) and several other chromatinopathies on Levi's residual differential with minimal incremental cost or blood volume.

This is the clearest single piece of external evidence for prioritizing the methylation-panel diagnostic item in Levi's workspace, and for separating the TBRS-residual component of the epigenetic overgrowth theory from the component that is tied to untested methylation / imprinting mechanisms.

Provenance