GRAF

GRAF (Genomic Analysis Visualization Tool) is an award-winning reference genome tool developed by Velsera, designed to overcome the fundamental limitations of canonical linear reference genomes. Traditional linear references are over 20 years old and fail to account for the full spectrum of genomic variation, leading to reference bias that disproportionately affects non-European populations. GRAF addresses this systemic blindness by leveraging genome graphs to enable fast, accurate, and comprehensive discovery of genetic variation in the human genome.

Velsera has been pioneering graph reference-based secondary NGS analysis technology since 2014, and GRAF represents the maturation of that work into robust, stable, scalable, and deeply validated pipelines ready for deployment in pharma discovery and clinical environments. As pan-genome-aware sequence analysis becomes the new standard — driven in part by the draft human pan-genome reference from the HPRC — GRAF positions users at the forefront of this methodological shift.

Core Technology and Approach

• Uses genome graphs to represent genetic variation across a population, with the canonical linear reference serving as the backbone and edges representing alternate sequences.
• Eliminates reference bias inherent in standard linear reference mapping, enabling more equitable and accurate analysis across diverse ancestries.
• Provides proprietary data structures, algorithms, and graph references that effectively capture the full breadth of genomic diversity.
• The graph aligner maps significantly more reads than conventional tools such as BWA, and accurately detects variants — for example, correctly identifying homozygous insertions that BWA erroneously calls as heterozygous.

Key Capabilities

• Improved alignment for compound variants: Graph alignment places variants within existing variants, facilitating more accurate alignment in complex genomic regions.
• Complex variant discovery: Facilitates identification of in-phase variants and other complex genomic events that are missed by linear reference approaches.
• Structural and point variant support: GRAF's solution spans both structural variants and single nucleotide variants, providing comprehensive coverage across variant types.
• Pan genome-aware secondary NGS analysis: Deeply validated pipelines are available for deployment in pharma discovery environments, backed by a robust technology and patent portfolio.

Who GRAF Is Designed For

• Clinical Diagnostic Laboratories: Enables commercialization of new diagnostic tests through custom-developed NGS analysis pipelines, with optional validation, regulatory support, Revenue Cycle Management, and tertiary CGW reporting.
• NGS Assay and Equipment Manufacturers: Supports bringing new tests and novel equipment to market, including pipeline creation, integration, hosting, and deployment, plus compliance and reimbursement support.
• Biotech and Pharma: Helps researchers cut through the noise of exploding clinical-genomic data volumes to accelerate discovery, supported by Velsera's in-house Bioinformatics and Science teams.
• Non-profits and Government Agencies: Widely adopted across government, academic, and commercial research settings, with Velsera's scientific teams available to help maximize the value of genomic assets.

Supported Solutions and Workflows

• Analysis and bioinformatics workflows for clinical and research genomics.
• NGS assay development and expansion, including custom pipeline creation, hosting, and deployment on the Seven Bridges platform.
• Integrated secondary and tertiary analysis packages available through optional CGW reporting.
• Validation and regulatory support to ensure compliance for laboratory customers.

GRAF is deployable in pharma discovery environments and integrates with Velsera's Seven Bridges platform. The technology is supported by a strong patent portfolio and has demonstrated real-world impact in large-scale genomics projects, including work aimed at capturing novel variants from underrepresented ancestries such as African and Native American populations within diverse genomic datasets.