Coursera Flash Sale
40% Off Coursera Plus for 3 Months!
Grab it
Explore a groundbreaking approach to error correction in DNA data storage through this 16-minute conference presentation that introduces Marker-GC+ (MGC+) codes for handling random edit errors. Learn how DNA storage faces significant challenges from synthesis, sequencing, and storage errors including deletions, insertions, and substitutions that compromise data reliability in short DNA sequences. Discover the limitations of current approaches that focus on specific error types rather than addressing all three simultaneously, and understand why existing DNA storage techniques primarily rely on sequencing redundancy rather than direct error correction. Examine how GC+ codes currently handle indels by mapping them to erasures and substitutions in higher fields using Reed-Solomon codes with guess-and-check decoding mechanisms. Master the novel MGC+ construction that enhances GC+ by inserting periodic markers and implementing a Maximum A Posteriori (MAP)-based decoding algorithm to estimate optimal offset patterns. Understand how this approach generates informed guesses rather than exhaustive searches, reducing Reed-Solomon redundancy requirements while improving overall code rates. Analyze comparative results showing MGC+ outperforming HEDGES and GC+ across information density, decoding latency under 2ms, and reliability metrics for 176-bit binary messages encoded into DNA under realistic error probabilities based on experimental data.
