Genetic Disorders
- Home
- Disciplines
- Genetic Disorders
Rare disease and inherited disorder analysis built for the most genetically diverse populations on earth — and the ones worst served by existing databases.
African populations carry more genetic variation than all other populations combined, and appear in reference databases such as gnomAD and ClinVar at a fraction of that diversity. The practical consequence is that a variant genuinely common and benign in a West African population may be absent from reference panels entirely, and get flagged as a rare candidate for disease.
This is the most common reason diagnostic pipelines return unmanageable candidate lists for African families. It is a solvable problem, and solving it is much of what we do in this area.
What We Analyse
From single families to national rare disease programmes.
Trio and Family Exome Analysis
De novo, autosomal recessive, X-linked and compound heterozygous filtering across trios and extended pedigrees, using inheritance models appropriate to the family structure actually present.
Consanguineous Pedigrees
Homozygosity mapping and runs-of-homozygosity analysis to narrow candidate regions — often far more powerful than frequency filtering alone.
Phenotype-Driven Prioritisation
Ranking candidates against HPO terms supplied by the clinical team using Exomiser and related tools, so the shortlist reflects the patient's actual presentation.
Variant Interpretation
Classification following ACMG-AMP guidelines, with explicit acknowledgement of where evidence is weak because the population is under-represented rather than because the variant is unimportant.
Sickle Cell and Haemoglobinopathies
Genotyping, modifier gene analysis, genotype-phenotype correlation and newborn screening programme data analysis.
Structural and Copy Number Variation
CNV and structural variant detection from short read, long read and array data, for the cases where exome sequencing returns nothing.
The Reference Data Problem, and What We Do About It
Population-appropriate frequency filtering
We filter against African-specific allele frequency resources — gnomAD African subsets, H3Africa data, and population-specific cohorts where available — rather than global frequencies dominated by European samples. This alone typically reduces a candidate list by an order of magnitude.
Honest reporting of annotation gaps
Where a gene or region has poor annotation coverage in African populations we say so in the report. A variant of uncertain significance in a well-studied gene, and one in a gene nobody has studied in your population, are not the same finding and should not be presented as though they were.
Reference bias in alignment
The standard human reference genome represents African haplotypes poorly, and reads from divergent regions align badly or not at all. Where the clinical question warrants it we use graph-based or pangenome references, and we flag regions where coverage gaps are a property of the reference rather than of the sample.
Contributing back
Every project of this kind adds to the evidence base. Where consent and ethics permit, we encourage and support deposition of variant and frequency data into public resources, because the next family's diagnosis depends on it.
