AI in Genomic Counseling

Variant Interpretation

ACMG Classification

class VariantInterpreter:
    def classify_variant(self, evidence_scores):
        total = sum(self.evidence_weights[k] * evidence_scores.get(k, 0)
                   for k in self.evidence_weights)
        if total > 0.9: return 'Pathogenic'
        elif total > 0.7: return 'Likely pathogenic'
        elif total > 0.3: return 'VUS'
        elif total > 0.1: return 'Likely benign'
        return 'Benign'

Risk Communication

class RiskCommunicator:
    def calculate_lifetime_risk(self, gene, variant, family_history, age):
        baseline = self.baseline.get(gene, 0.1)
        or_value = self._get_odds_ratio(gene, variant)
        family_mod = self._compute_family_modifier(family_history)
        return min(baseline * or_value * family_mod, 0.95)

Family History Analysis