M-DAB: An Input-Distribution Optimization Algorithm for Composite DNA Storage by the Multinomial Channel

TL;DR

This paper introduces M-DAB, an optimization algorithm for maximizing capacity in composite DNA storage channels modeled as multinomial channels, demonstrating improved input distribution design for DNA data storage.

Contribution

It proposes the M-DAB algorithm, a generalized dynamic assignment method for optimizing input distributions in multinomial DNA storage channels, extending previous binomial channel approaches.

Findings

Capacity scales with support size of the optimal input distribution.

M-DAB effectively finds capacity-achieving distributions for complex DNA channels.

Empirical results confirm the algorithm's effectiveness across different read scenarios.

Abstract

Recent experiments have shown that the capacity of DNA storage systems may be significantly increased by synthesizing composite DNA letters. In this work, we model a DNA storage channel with composite inputs as a \textit{multinomial channel}, and propose an optimization algorithm for its capacity achieving input distribution, for an arbitrary number of output reads. The algorithm is termed multidimensional dynamic assignment Blahut-Arimoto (M-DAB), and is a generalized version of the DAB algorithm, proposed by Wesel et al. developed for the binomial channel. We also empirically observe a scaling law behavior of the capacity as a function of the support size of the capacity-achieving input distribution.