In Development

Next-Generation Prenatal Screening
In Development

We are building an AI foundation model targeting genetic condition detection at foetal fractions as low as 1-2% — aiming to enable earlier testing and reduce the test failures that delay critical medical decisions.

Why Current NIPT Falls Short

Current non-invasive prenatal testing (NIPT) fails 3-8% of the time due to low foetal DNA concentration in maternal blood. When tests fail, families face weeks of anxious waiting for repeat testing. Testing cannot reliably happen before 10 weeks of pregnancy, limiting options for early intervention.

Our AI foundation model is designed to detect patterns in cell-free DNA that are invisible to the statistical methods developed in 2008. By learning the complex relationships between fragment sizes, methylation patterns, and genomic coverage, we aim to transform what would be test failures into reliable results — and enable testing 3-4 weeks earlier than currently possible.

6-7Weeks — target gestation
50-70%Projected failure reduction
107Conditions targeted
1-2%Target foetal fraction

Technical Features

Our AI platform is being designed to deliver clinical-grade genetic screening with improved sensitivity and accuracy.

Ultra-Low Foetal Fraction Analysis

Targeting detection at foetal fractions as low as 1-2%, compared to the industry standard requirement of 4%+. This would enable testing from 6-7 weeks gestation, versus the current standard of 10+ weeks.

1-2%Target foetal fraction
6-7 wkTarget gestation

Population-Specific Optimisation

Ancestry-aware analysis designed to reduce algorithmic bias affecting non-European populations. Our AI models will be trained across diverse genomic datasets to improve equitable accuracy for all families.

GlobalPopulation coverage goal
ReducedAlgorithmic bias target

Comprehensive Condition Screening

Designed to screen for 107 genetic conditions in a single test, including common trisomies, sex chromosome aneuploidies, microdeletions, microduplications, monogenic disorders, oncology hotspots, and repeat expansions.

107Conditions targeted
1Test required

Real-Time Quality Assessment

Planned automated confidence scoring for every result with transparent uncertainty quantification. The AI will distinguish between biological limitations and technical issues, providing clinicians with actionable insights.

100%Results with confidence scores
TransparentAI explainability

Platform-Agnostic Design

EabhaSeq is designed to integrate with existing laboratory infrastructure. The platform will work with standard sequencing equipment and sample collection protocols, requiring no hardware replacement or workflow disruption.

SequencingIllumina and Nanopore compatible
IntegrationAPI integration with existing LIMS
ProtocolsStandard maternal blood draw
HardwareNo replacement required

Clinical Validation Programme

EabhaSeq's prenatal platform is currently in development and has not yet been clinically validated. Our synthetic data generation technology, which powers the training pipeline, is validated and commercially available today.

We are accepting partnership enquiries from diagnostic laboratories, healthcare systems, and clinical research sites interested in participating in our validation programme.

Transform Prenatal Care

Join us in making genetic screening earlier, more accurate, and accessible to all families worldwide.