Fabric GEM

Fabric GEM is an AI-driven genomic interpretation platform developed by Fabric Genomics, designed to enable accurate and near-instant identification of disease-causing genes in whole genome sequencing (WGS) and whole exome sequencing (WES) workflows. It is built for clinical laboratories, lab directors, and physicians who need to scale next-generation sequencing (NGS) diagnostics without sacrificing accuracy or increasing review burden.

By leveraging proprietary algorithms, the largest public and proprietary datasets, and intelligent integration with phenotypic data, Fabric GEM prioritizes the disease-causing variant at the top of a ranked candidate list. Validated at multiple collaborator centers including Rady Children's Institute for Genomic Medicine, the platform reduces average clinical review time to approximately 15 minutes and enables labs to handle up to 10 times more cases with the same team.

Key Performance Metrics

• 98% of causal variants ranked in the top 5 candidates in retrospective studies
• 90% of causal genes ranked 1st or 2nd in pediatric case studies
• 100% of causal variants ranked in the top 10 in validation studies
• Average clinical review reduced to just 2 genes per case for single proband cases
• 5–10% new diagnoses achieved through re-analysis of previously negative cases
• Clinical review time reduced by more than 90% compared to traditional workflows

Core Capabilities and Features

• Advanced AI ranking: Multi-dimensional analysis of all variants to surface the most likely disease-causing candidate at the top of the list, reducing typical candidate review from 20–50 down to fewer than five
• Automated deep phenotyping: Integrates phenotypic data to deliver more accurate genomic interpretation without manual curation overhead
• Comprehensive variant evaluation: Assesses both small variants and structural variants (SVs) within a single workflow
• Artifact and ancestry detection: Identifies common sequencing artifacts and cryptic ancestry to improve result reliability
• Consanguinity and inheritance prediction: Incorporates Truploidy, a Fabric AI algorithm that predicts consanguinity based on analysis of segments of homozygosity and adjusts for penetrance based on data
• Knowledge synthesis: Combines agnostic variant search with clinical knowledge sources including OMIM, ClinVar, and condition-gene databases, building on Fabric's foundational VAAST and Phevor algorithms
• Data transparency: Provides clear supporting evidence alongside ranked candidates to support confident clinical decision-making

Validated Use Cases

• Rare and undiagnosed diseases
• Ultra-rapid NICU and PICU cases where time to diagnosis is critical
• Re-analysis of previously negative cases to uncover new diagnoses
• Solo proband and nuclear family testing configurations

Clinical Impact and Workflow Benefits

• Addresses the primary bottleneck in clinical genomics workflows — the review and evaluation of candidate variants — which can otherwise consume 24 hours or more per case using traditional filtering techniques
• Enables labs to confidently scale testing volume by dramatically reducing per-case review time
• Particularly impactful for time-sensitive pediatric conditions in the NICU, where delays in diagnosis can affect treatment outcomes and quality of life
• Validated with hundreds of cases across multiple clinical collaborator centers, including Rady Children's Institute for Genomic Medicine

Fabric GEM has been validated in multiple clinical settings for WES and WGS testing, demonstrating broad utility across rare disease diagnostics. The platform is designed to integrate into existing clinical lab workflows, empowering teams to scale NGS diagnostics with automation, speed, and accuracy.