Folklore

Folklore is a variant classification platform for clinical genetics laboratories, developed by Helena Bioinformatics. It takes a whole-genome or whole-exome VCF as input and delivers classified, prioritised variants ready for specialist review in under 10 minutes. The platform is designed to support routine laboratory work across panels, exomes, and whole genomes, bringing nuclear variants, mitochondrial DNA, structural variants, phenotype data, inheritance patterns, clinical screening signals, and literature evidence together into a single integrated case record.

The end-to-end workflow moves from VCF ingest through annotation, classification, prioritisation, geneticist review, and final signed report. Annotation draws on 45 curated databases covering allele frequency, conservation scores, and computational predictions. Classification applies deterministic rule sets — ACMG/AMP criteria for nuclear variants, the MMDWG 2020 framework for mitochondrial DNA, and Riggs 2020 criteria for structural variants. Prioritisation incorporates phenotype, screening, and cohort signals to order the gene list before a geneticist inspects the fired criteria and signs off. Every report is tiered and traceable, linking each result back to its supporting evidence.

Modules

• Variant Analysis & Classification — Classifies millions of variants into five ACMG classes and ranks them by priority. Versioned at V3.39.1.
• Mitochondrial DNA Analysis — Applies rule-based classification under the MMDWG 2020 framework for mtDNA variants.
• Structural Variant Analysis — Handles deletions and duplications using Riggs 2020 criteria, assigning one of five tiers.
• Phenotype Matching — Maps HPO terms to clinical priority on a scored scale of 0–100 across five tiers.
• Clinical Screening — Organises variants by review order rather than pathogenicity, using a Tier 1–4 system.
• Cohort Analysis — Analyses groups of samples for burden, enrichment, and candidate variant identification using Fisher-based statistics.
• Family & Trio Analysis — Supports child-mother-father trio configurations to determine the origin of a variant, including de novo calling.
• Literature Evidence — Links variants to relevant, anchored publications sourced from PubMed.

Performance and Architecture

• Whole-genome classification and prioritisation completes in under 10 minutes through locally indexed reference data and parallel processing of independent analytical stages.
• Framework-specific classifiers for nuclear, mitochondrial, and structural variant types run within a single unified pipeline.
• Classification and patient-specific prioritisation are delivered together in one review workspace.

AI-Native Interpretation

• An AI assistant is available throughout the Folklore workflow and operates from the full case context already assembled — including classifications, phenotype data, inheritance information, quality metrics, and linked literature.
• The assistant can surface and explain significant findings, such as identifying a pathogenic variant and summarising its associated condition and inheritance pattern, to support the reviewing geneticist.
• Final sign-off remains with a qualified specialist; the assistant functions as decision support, not a diagnostic device.

Clinical Infrastructure and Compliance

• EU data residency — Patient data is stored in Helsinki, Finland and does not leave the EU.
• Encryption — Data is encrypted both in transit and at rest.
• Role-based access — Access controls are applied at the user role level.
• Versioned evidence — Each classification result records the classifier version and the supporting evidence used at the time of analysis.
• Public methodology — Classification criteria, decision thresholds, and rules are openly documented.
• Qualified sign-off — Every final report requires review and signature by a specialist.
• Folklore is positioned as decision support and is not a CE-marked in vitro diagnostic device.

Folklore supports panels, exomes, and whole genomes within the same workflow, making it applicable across a range of clinical genomics use cases from routine screening to complex rare disease investigation.