A Sequence Obfuscation Method for Protecting Personal Genomic Privacy

Area of Science:

• Genomics
• Bioinformatics
• Computational Biology

Background:

• Advances in whole genome sequencing technology have made large-scale personal genomic data collection feasible and affordable.
• This surge in genomic data collection for medical diagnoses and drug discovery presents significant challenges for personal genomic privacy protection.
• Current genomic privacy methods often suffer from time-consuming encryption or low data recovery accuracy.

Purpose of the Study:

• To develop a novel method for protecting personal genomic privacy that is both fast and reliable.
• To ensure that personal genomic data remains utilizable for research and medical applications while maintaining confidentiality.
• To address the limitations of existing genomic privacy-protection techniques.

Main Methods:

• Proposes IterMegaBLAST, a sequence similarity-based obfuscation method for personal genomic privacy.
• Utilizes MegaBLAST to identify the most similar sequence to a randomly selected sequence within a dataset.
• Clusters aligned sequences and generates obfuscated sequences using a DNA generalization lattice scheme iteratively.

Main Results:

• IterMegaBLAST significantly outperforms existing state-of-the-art approaches.
• Demonstrates superior performance in both utility accuracy and time complexity under equivalent anonymity levels.
• Experimental results on benchmark datasets validate the method's effectiveness.

Conclusions:

• IterMegaBLAST provides an efficient and effective solution for safeguarding personal genomic privacy.
• The method successfully balances the need for genomic data utility with robust privacy protection.
• This approach is crucial for the responsible advancement of genomic medicine and research.