Code Design for Duplex Read Sequencing

Motivated by duplex read sequencing developed by Oxford Nanopore Technologies, this paper proposes a concatenated coding scheme for nanopore sequencers where DNA sequences are decoded from noisy reads of the template and reverse-complement strands from the same DNA molecule, that is, a duplex read. First, we show that the double-strand pairwise error probability (PEP) bound is multiplicative with respect to the single-strand PEP bounds of the template and reverse-complement codebooks, thus giving duplex decoders significantly lower error rates compared to simplex decoders which only use the template strand. Then, we propose a decoder for multiple concatenations of short codebooks designed using the double-strand PEP bound.