Work

Synthetic Germline VCF Generation for Rare Diseases: Case Study in NF1

Year: 2026

Type: conference-paper

Abstract: Synthetic data generation (SDG) holds promise for rare disease research by enabling privacy-conscious data sharing and AI model training on limited cohorts. We present a framework for generating synthetic germline variant call format (VCF) datasets for Neurofibromatosis type 1 (NF1), combining domain knowledge with a CART-based generator trained on whole exome sequencing data from 182 NF1 patients. Three progressively refined generation strategies are evaluated across fidelity, utility, and privacy dimensions. Our results demonstrate that realistic synthetic germline VCF cohorts can be generated for rare disease research, with privacy–fidelity tradeoffs motivating future work on formal privacy guarantees.

Source: ACM International Conference on Bioinformatics, Computational Biology, and Health Informatics (ACM-BCB)

DOI: 10.1145/3807503.3816791

Authors: Sikha Pentyala, Ziwei Pan, Patrick J McKeever, Jineta Banerjee, Luca Foschini, Martine De Cock

Institutions: University of Washington Tacoma, Sage Bionetworks

Language: English

Topic: Genomics and Rare Diseases

Subfield: Genetics

Field: Biochemistry, Genetics and Molecular Biology

Domain: Life Sciences

SDG: Good health and well-being

Open Access status: hybrid

Funder: National Science Foundation