Article
1. Introduction
A study on the ALPL gene, presented by GenoGra at BioTechX Europe 2026 in Basel, explores how a locus-specific pangenome can provide a broader representation of genomic variation and an integrated framework for investigating the evidence used in variant interpretation in hypophosphatasia.
Hypophosphatasia is a rare inherited disorder caused by variants in ALPL, with a clinical spectrum ranging from severe perinatal disease to milder skeletal and dental manifestations in adulthood.
2. The Challenge
Variant analysis still commonly relies on alignment against a single linear reference genome. While this remains the established approach, a single reference represents only one version of the human genome and may provide an incomplete view of variation across a locus, particularly for complex and structural variants.
Representation is only part of the challenge. Evidence used to interpret variants is distributed across different databases, where annotations and pathogenicity classifications do not always agree. Bringing genomic variation and its associated evidence into the same analytical context is therefore critical for more informed interpretation.
3. The GenoGra Approach
Using the GenoGra platform, we constructed a locus-specific pangenome of ALPL, combining the GRCh38 reference with variants from multiple VCF resources.
Instead of representing variants only as deviations from a single reference, the graph preserves alternative alleles as interconnected paths, providing a broader representation of variation across the locus.
The pangenome was enriched with genotypic, phenotypic and clinical annotations, enabling genomic variation to be explored together with the evidence available for its interpretation. Within the GenoGra platform, these different layers can be explored alongside graph topology, alternative paths and variant distributions. Information from different sources was also harmonized to allow direct comparison of annotations and pathogenicity classifications while preserving the evidence reported by each resource.
4. Key Results
The resulting ALPL pangenome encodes 4,589 variants: 3,897 SNPs, 670 small and complex variants and 22 structural variants. Although structural variants represent only a small fraction of the catalogue, the graph retains them as alternative paths rather than reducing them to differences against a single reference.
As additional haplotypes were incorporated, the pangenome grew only marginally, with its total sequence stabilising at approximately 114.5 kbp and its strict core at approximately 103.5 kbp. The remaining accessory sequence represents rare variation absent from the linear reference but retained within the graph.
Cross-database analysis also identified discordant pathogenicity classifications for 27.2% of the analysed variants. By preserving rather than collapsing these classifications, the harmonized framework makes it possible to compare conflicting evidence directly alongside the genomic and functional information associated with each variant.
5. Conclusion
The ALPL study shows how pangenomics can extend beyond a broader representation of genomic diversity.
By combining different classes of variation with genomic, functional and clinical evidence, a locus-specific pangenome provides a common framework in which variants and the information used to interpret them can be investigated together.
The same approach can be extended to other disease-associated genes and progressively enriched with additional genomic and clinical evidence.
In this perspective, the pangenome is not only a map of genomic diversity, but a framework for reading it in the light of the available evidence.