Oxford Nanopore Technologies (ONT) and Fabric Genomics collaborated to develop a scalable software solution for the end-to-end analysis and clinical reporting of nanopore sequencing data.

The initial applications will focus on pediatric and neonatal intensive care (NICU/PICU), where the need is acute and time matters most. In this setting, the platform supports the characterisation of rare disease in babies and children, helping to reduce the Diagnostic Odyssey and improve patient lives.

Oxford Nanopore - Fabric Genomics Integration

Fabric Enterprises’ GEM is the AI engine that powers the candidate scoring and prioritization.

Model-based: GEM is not a filtering pipeline. GEM employs a probabilistic model — trained on thousands of samples — that integrates multiple sources of clinical and genomic evidence to generate a Bayes factor score.
Transparent: GEM is not a black box. GEM provides an explanation for its candidate variants with respect to inheritance mode, penetrance, sequence quality, phenotype, consanguinity, ancestry, loss of function and other characteristics
Platform specific: GEM uses an AI-model parameterized for the ONT long-read sequencing platform, optimized on more than 100 genomes to maximize accuracy and reduce false positives, especially for SVs.
Validated: GEM has been rigorously evaluated against 100s of clinically confirmed samples representative of the real diversity of genetic diseases and variant types

To learn more about GEM3 and find out more about how it works, please contact us via the form above.