A memorable musical number from Cats muses on the loss of “Memory;” however, an unforgettable new feline-related study now describes the encoding of images of tigers (and pandas!) using printed epigenetic bits (“epi-bits”) on DNA as a novel storage approach for our data and dearest recollections. Can epi-bits make any memory loss a thing of the past?
An unforgettable team directed by Cheng Zhang, Qi Ouyang, Long Qian (Peking University), and Hao Yan (Arizona State University) understood that the density, durability, and low energy consumption associated with DNA-based memory storage could support the ever-expanding global “datasphere”; however, encoding data into nucleotide sequences via de novo synthesis suffers from significant limitations. Their new loss-less study took inspiration from epigenomic inheritance and synthetic DNA self-assembly and brought forth an unconventional memory storage framework that writes epigenetic modifications (5-methylcytosine; 5mC) onto DNA as “epi-bits.” This fascinating strategy supports synthesis-free parallel DNA data storage and high-throughput retrieval via nanopore sequencing and specifically designed processes/algorithms.
Let’s hear more about how “epi-bit” encoded images of tigers and pandas reveal the future of DNA memory storage:
- Writing epi-bits involves transferring non-methylated/methylated CpGs (0/1) by DNMT1 and S-adenosylmethionine from prefabricated single-stranded (ss)DNA bricks (“movable type”) assembled with ssDNA carriers onto a template
- This DNA self-assembly-based “typesetting” strategy supports the encoding of the ASCII code for the letters ‘DNA’ (24 bits) into DNA and the retrieval of the message with negligible background noise
- Scaled-up epi-bit writing employs barcoded DNA carriers, a universal template, and movable types that support 800-bit data encoding (structure drawings of DNA bases and modified derivatives) and data retrieval with a low error rate
- Even larger-scale data storage with multiple longer DNA templates and denser epi-bit sites on ssDNA bricks achieves 350-bit parallelism (the number of bits written in a single reaction per data-writing cycle)
- This approach supports the DNA storage of a ~17,000-bit image of a tiger with 96% accuracy after the implementation of several advanced data retrieval methodologies and algorithms after nanopore reading
- Integrating all strategies to store a larger ~250,000-bit panda image supports perfect image restoration
- iDNAdrive – a distributed, custom-made DNA storage experiment – allows users to write personal data into DNA
- iDNAdrive permits 60 student volunteers of diverse academic backgrounds to manually store pieces of text (5,000 bits) in epi-bits through a simple writing kit in a classroom
- The high accuracy of subsequent nanopore reads underpins the robustness of data writing, with 12/15 texts successfully restored and securely returned to each individual via an online server
Cats, tigers, pandas, and innovative epigenetics – this remarkable and (hopefully!) unforgettable study reports parallel, programmable, stable, and scalable DNA data storage as epi-bits as the solution to safely and securely safeguarding our over-abundance of papers, figures, songs, movies, and – most importantly – precious memories.
For more on how storing epi-bit images of tigers and pandas on DNA may stop memory loss, see Nature, October 2024.