Modeling DNA storage retrieval reliability via sequencing coverage depth.

DNA storage offers exceptional durability and density but remains constrained by high sequencing cost and limited throughput. Sequencing coverage depth is a key determinant of retrieval reliability and sequencing efficiency, yet its quantitative characterization requires realistic modeling of channel nonuniformity beyond idealized uniform assumptions. In this work, we develop a nonuniform coverage-depth analysis framework for DNA storage channels by modeling polymerase chain reaction and sequencing data, with a log-normal instantiation for the empirical channel distribution. The proposed framework covers both nonuniform noiseless and noisy channels and is used to analyze the MDS coverage depth problem. For the noiseless case, we derive the expected minimum sequencing coverage depth required for complete decoding under a nonuniform channel distribution. For the noisy case, we present two complementary theories: a coupon-collector-problem-based (CCP) theory that provides a conservative probabilistic envelope, and a critical-quantile-based theory (CQT) that refines the estimate by identifying the key coverage-depth range that determines whether most strands are recovered. We further compare the uniform-channel theory, CCP theory, and CQT theory across multiple code rates and retrieval parameters through extensive Monte Carlo simulations. To support practical use, we also develop an interactive web platform for parameter tuning and visualization. The resulting theory and platform provide theoretical foundations and practical tools for improving retrieval reliability and sequencing efficiency in DNA storage systems.