R&D: DNA-Aeon Provides Flexible Arithmetic Coding for Constraint Adherence and Error Correction in DNA Storage

Nature Communications has published an article written by Marius Welzel, Peter Michael Schwarz, Hannah F. Löchel, Department of Mathematics and Computer Science, University of Marburg, Marburg, Germany, and Center for Synthetic Microbiology (SYNMIKRO), University of Marburg, Marburg, Germany,Tolganay Kabdullayeva, Center for Synthetic Microbiology (SYNMIKRO), University of Marburg, Marburg, Germany, Sandra Clemens, Department of Mathematics and Computer Science, University of Marburg, Marburg, Germany, and Center for Synthetic Microbiology (SYNMIKRO), University of Marburg, Marburg, Germany, Anke Becker, Center for Synthetic Microbiology (SYNMIKRO), University of Marburg, Marburg, Germany, Bernd Freisleben, and Dominik Heider, Department of Mathematics and Computer Science, University of Marburg, Marburg, Germany, and Center for Synthetic Microbiology (SYNMIKRO), University of Marburg, Marburg, Germany.

Abstract: “The extensive information capacity of DNA, coupled with decreasing costs for DNA synthesis and sequencing, makes DNA an attractive alternative to traditional data storage. The processes of writing, storing, and reading DNA exhibit specific error profiles and constraints DNA sequences have to adhere to. We present DNA-Aeon, a concatenated coding scheme for DNA data storage. It supports the generation of variable-sized encoded sequences with a user-defined Guanine-Cytosine (GC) content, homopolymer length limitation, and the avoidance of undesired motifs. It further enables users to provide custom codebooks adhering to further constraints. DNA-Aeon can correct substitution errors, insertions, deletions, and the loss of whole DNA strands. Comparisons with other codes show better error-correction capabilities of DNA-Aeon at similar redundancy levels with decreased DNA synthesis costs. In-vitro tests indicate high reliability of DNA-Aeon even in the case of skewed sequencing read distributions and high read-dropout.“