Abstract: An embodiment may involve obtaining an image composed of color channels, wherein each of the color channels represents colors therein with n bits per pixel; quantizing each of the color channels to be represented by m bits per pixel; using a space-filling curve to convert each of the color channels as quantized into respective vectors; generating, for each of the respective vectors, a respective set of m arrays, wherein the m arrays for a particular vector of the respective vectors contain indices of where a fixed value selected from a range of 0 to m?1 is found in the particular vector; applying, to each of the m arrays for each of the respective vectors, differential encoding and compression of the indices therein; mapping each of the m arrays for each of the respective vectors to blocks of nucleotides; synthesizing the blocks; and storing, in a DNA-based storage medium, the blocks.

Abstract: An embodiment may involve obtaining an image composed of color channels, wherein each of the color channels represents colors therein with n bits per pixel; quantizing each of the color channels to be represented by m bits per pixel; using a space-filling curve to convert each of the color channels as quantized into respective vectors; generating, for each of the respective vectors, a respective set of m arrays, wherein the m arrays for a particular vector of the respective vectors contain indices of where a fixed value selected from a range of 0 to m-1 is found in the particular vector; applying, to each of the m arrays for each of the respective vectors, differential encoding and compression of the indices therein; mapping each of the m arrays for each of the respective vectors to blocks of nucleotides; synthesizing the blocks; and storing, in a DNA-based storage medium, the blocks.

Abstract: The disclosure relates to a re-writable DNA-based digital storage system with a random access feature. An example embodiment includes selecting address representations of m nucleotide sequences of n bases each. Each of the address representations of m nucleotide sequences (i) consists of approximately 50% guanine and cytosine content and (ii) is self-uncorrelated. The address representations are mutually uncorrelated with one another. All of the address representations end with a particular base. The embodiment also includes selecting, for a particular address representation from the address representations of the m nucleotide sequences, a corresponding data representation of a nucleotide sequence of L bases. The embodiment further includes concatenating the particular address representation with the data representation to form a representation of a target nucleotide sequence of n+L bases. In addition, the embodiment includes synthesizing the target nucleotide sequence.

Abstract: The disclosure relates to a re-writable DNA-based digital storage system with a random access feature. An example embodiment includes reading a physical nucleotide sequence of 2n+L bases to form a digital representation of the physical nucleotide sequence; dividing the digital representation of the physical nucleotide sequence into an address representation of n bases, followed by a data representation of L bases, followed by a further address representation of n bases; decoding the data representation into an integer value less than 3L that is a sum of a first addend and a second addend, wherein the data representation includes a first subsequence of bases encoding the first addend followed by a second subsequence of bases encoding the second addend; and storing, in a computer memory, the integer value.

Abstract: The present disclosure provides DNA-based storage system demonstrated through experimental and theoretical verification that such a platform can easily be implemented in practice using portable, nanopore-based sequencers. The gist of the approach is to design an integrated pipeline that encodes data to avoid synthesis and sequencing errors, enables random access through addressing, and leverages efficient portable nanopore sequencing via new anchored iterative alignment and insertion/deletion error-correcting codes. The embodiments herein represent the only known random access DNA-based data storage system that uses error-prone portable, nanopore-based sequencers and produces low-error readouts with the highest reported information rate and density.