# The power of mouse models in the diagnostic odyssey of patients with rare congenital anomalies

## TL;DR

Mouse models help diagnose rare birth defects by linking genetic variants to their effects, improving understanding and reducing diagnostic delays.

## Contribution

The paper highlights advances in using engineered mice to bridge variant discovery and diagnosis in congenital anomalies.

## Key findings

- Engineered mice enable validation of gene function and disease mechanisms in congenital anomalies.
- Advances in genomics and genome editing improve the creation of clinically relevant animal models.
- Integration of animal models with patient data is crucial for resolving molecular diagnoses.

## Abstract

Congenital anomalies are structural or functional abnormalities present at birth, which can be caused by genetic or environmental influences. The availability of genome sequencing has significantly increased our understanding of congenital anomalies, but linking variant identification to functional relevance and definitive diagnosis remains challenging. Many genes have unknown or poorly understood functions, and with a lack of clear genotype-to-phenotype correlations, it can be difficult to move from variant discovery to diagnosis. Thus, for most congenital anomalies, there still exists a “diagnostic odyssey” which presents a significant burden to patients, families and society. Animal models are essential in the gene discovery process because they allow researchers to validate candidate gene function and disease progression within intact organisms. However, use of advanced model systems continues to be limited due to the complexity of efficiently generating clinically relevant animals. Here we focus on the use of precisely engineered mice in variant-to-function studies for resolving molecular diagnoses and creating powerful preclinical models for congenital anomalies, covering advances in genomics, genome editing and phenotyping approaches as well as the necessity for future initiatives aligning animal modelling to deep patient multimodal datasets.