Rewritable Two-Dimensional DNA-Based Data Storage with Machine Learning Reconstruction

Abstract

DNA-based data storage platforms traditionally encode information only in the nucleotide sequence of the molecule. Here, we report on a two-dimensional molecular data storage system that records information in both the sequence and the backbone structure of DNA. Our "2DDNA" method efficiently stores high-density images in synthetic DNA and embeds metadata as nicks in the DNA backbone. To avoid costly redundancy used to combat sequencing errors and missing information content that typically requires additional synthesis, specialized machine learning methods are developed for automatic discoloration detection and image inpainting. The 2DDNA platform is experimentally tested on a library of images that show undetectable visual degradation after processing, while the image metadata is erased and rewritten to modify copyright information. Our results show that DNA can serve both as a write-once and rewritable memory for heterogenous data. Moreover, the storage density of the molecules can be increased by using different encoding dimensions and avoiding error-correction redundancy.

On the prediction of arginine glycation using artificial neural networks

“…While artificial intelligence (AI) is the overarching science of mimicking human abilities, machine learning is a specific subset of AI that trains a machine how to learn. Nowadays, machine learning is one of the most important tools for scientists in the development of new applications 43,44 . We could have conveniently employed a linear regression-based approach to estimate the probability of glycation, but the results would have been considerably poor compared to those obtained using the more sophisticated ANN.…”

Molecular Encryption and Steganography Using Mixtures of Simultaneously Sequenced, Sequence-Defined Oligourethanes

“…These macromolecules, often referred to as digital polymers, store information at the molecular level in the form of a defined and absolute monomer sequence (i.e., primary structure). − Encoding information at the molecular level can be used to surmount some drawbacks of conventional storage devices, such as durability, longevity, and excessive spatial occupation . Such macromolecular information storage is now well established with artificial DNA biopolymers, which have been shown to store and retrieve significant amounts of information. − Two strengths of using DNA for information storage are the ability to replicate and retrieve data, as well as the ability to exploit rapid advances in sequencing, such as Next-Gen methods , and nanopore technology. − Alongside DNA, advances in the synthesis and sequencing of abiotic SDPs have improved the information storage capabilities of these macromolecular systems, − with commensurate advances toward this goal seen with multicomponent reactions − and small molecule strategies. − However, significant advances are still needed to rival the effective storage capacity of nucleic acids using SDPs, let alone silicon-based data storage. , …”

The Input and Output Entropies of the $k$-Deletion/Insertion Channel

“…In the last decade, channels that introduce insertion and deletion errors attracted significant attention due to their relevance to DNA storage systems [3], [19], [33], [34], [44], [48], where deletion and insertion are among the most dominant errors [23], [38]. The study of communication channels with insertion and deletion errors is also relevant to many other applications such as the synchronization of files and symbols of data streams [14] and for cases of over-sampling and under-sampling at the receiver side [39].…”