CIF: NSF-BSF: Small: Collaborative Research: Characterization and Mitigation of Noise in a Live DNA Storage Channel

University Of Virginia Main Campus, Charlottesville VA

Investigators

Abstract

Deoxyribonucleic acid (DNA) is emerging as a potential solution to growing data storage needs, due to recent advances in DNA synthesis and sequencing. Compared to silicon-based storage, DNA offers higher density and data longevity. Data-bearing DNA can be stored in non-biological environments (in vitro) or be embedded into the DNA of living organisms such as bacteria or plants (in vivo). In vivo DNA data storage benefits from a natural protective shell and a reliable and cost-efficient way to copy data. Furthermore, the compatibility of in vivo storage with living cells creates a potential for advances in synthetic-biology algorithms that require a mechanism for data storage. The goal of this project is to develop and demonstrate an in vivo DNA data storage system that is resilient to errors arising during sequencing and synthesis and from mutations, while achieving the maximum possible data density. To achieve its goal, the project relies on the following research thrusts: 1) Characterization of the live DNA channel and 2) Error mitigation and correction. Errors in DNA storage are known to be context-dependent, i.e., they depend on the local structure of the sequence. To enable error control, a mathematical model for the channel of live DNA storage will be formulated and the dependence of error rates on sequence context will be characterized. This project will take advantage of new advances in gene-editing technologies, in particular the CRISPR/Cas system, to characterize the channel through experiments. The second thrust aims to overcome the limitations of the existing error-control techniques, which either do not provide adequate error protection or are highly redundant. The project will develop semiconstrained-coding techniques for limiting the frequency of patterns with high error rates and high-rate error-correction schemes for combating errors. The semiconstrained and error-correction methods will be adapted to the live DNA channel and then implemented to evaluate their performance in practice and to demonstrate their advantages. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.