Error-correction codes for data storage on synthetic DNA molecules
| When |
Jan 26, 2026
from 11:00 to 12:00 |
|---|---|
| Where | Salle Condorcet |
| Contact Name | Nelly Pustelnik |
| Attendees |
Elsa Dupraz |
| Add event to calendar |
vCal
iCal
|
DNA-based data storage is an emerging technology that offers significantly higher density and durability compared to conventional storage systems (hard drives, tapes, etc.). It consists of encoding binary information into synthetic DNA molecules composed of nucleotides A,C,G, T, using a process called synthesis. To retrieve the original data, the DNA molecules are read using a biotechnological process called sequencing. However, both the DNA storage medium and the synthesis/sequencing processes introduce substantial error rates (ranging from 2% to 5%) in the read data. These errors include not only substitutions, but also insertions and deletions which cannot be corrected by conventional error-correction codes originally designed for telecommunications. This limitation necessitates the development of novel specific error-correction techniques. A key opportunity when designing these techniques lies in the fact that the sequencing outputs a large number of reads of the same molecule, each with different error patterns.
In this presentation, we first introduce the main principles of DNA data storage. We then present novel error-correction approaches capable of addressing all three error types. We show that the most effective error-correction solutions combine consensus algorithms from bioinformatics, designed to efficiently process multiple reads, with advanced error-correction schemes to mitigate residual insertions, deletions, and substitutions. Finally, we discuss key challenges that are currently addressed to scale this technology effectively. This field of research lies at the intersection of signal processing, bioinformatics, and biology, making it a fascinating and multidisciplinary topic.