Coding For Dna Storage With Nanopore Sequencing
summary
Uses Oxford Nanopore nanopore reads and the Scrappie emulator to model squiggles as noisy, randomly stretched current traces with pore memory moving from 5 to 9 bases (≫250k states). Designs a 256-codeword squiggle codebook on 100–130 base strands, concatenated with detectable comma spacers, decodes via DTW pairwise metrics or a tiny 3‑layer CNN, and reports about 20–30% single‑read success rising to reliable identification after a few reads; SNR ≈5–10 dB and practical GPU‑free, portable decoding.
AI-written summary · about
chapters (9)
- 0:00DNA storage research intro
- 5:31DNA as communication model
- 10:58Nanopore sequencing explained
- 20:42Squiggle coding and soft decoding
- 31:04Channel modeling challenges
- 35:28Codebook design for squiggles
- 43:12Performance results and future
- 45:45Neural decoding approaches
- 48:15Q&A
related talks
10mRevisiting Network Coding for Warm Blob Storage
31mMulti-version Coding and Shared Memory Emulation
01h 14mErasure Coding at Scale
57mCoding for Interactive Communication
11mExploiting Combined Locality for Wide-Stripe Erasure Coding in Distributed Storage
17mParaRC: Embracing Sub-Packetization for Repair Parallelization in MSR-Coded Storage
15mDistServe: Disaggregating Prefill and Decoding for Goodput-optimized Large Language...
12msub0 Asia 2024 - Axay Sagathiya (Chainsafe)