Cited by 173DNA and Biological ComputingAlgorithms and Data CompressionAdvanced Data Storage Technologies

The DOI matches its Crossref record; 28 of the 30 references it cites were found in a registry or an index.

Low cost DNA data storage using photolithographic synthesis and advanced information reconstruction and error correction

Open access

Abstract

Due to its longevity and enormous information density, DNA is an attractive medium for archival storage. The current hamstring of DNA data storage systems—both in cost and speed—is synthesis. The key idea for breaking this bottleneck pursued in this work is to move beyond the low-error and expensive synthesis employed almost exclusively in today’s systems, towards cheaper, potentially faster, but high-error synthesis technologies. Here, we demonstrate a DNA storage system that relies on massively parallel light-directed synthesis, which is considerably cheaper than conventional solid-phase synthesis. However, this technology has a high sequence error rate when optimized for speed. We demonstrate that even in this high-error regime, reliable storage of information is possible, by developing a pipeline of algorithms for encoding and reconstruction of the information. In our experiments, we store a file containing sheet music of Mozart, and show perfect data recovery from low synthesis fidelity DNA.

Overview

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1038/s41467-020-19148-3

Cites

28 of 28 references verified0 citation errors · 0 not checkable

Statistics

Figures

Fig. 1 Encoding scheme for storing 100 kB of music data in 16,383 DNA sequences. The stored information is structured as 14 packages containing 6 bits (L) with a length of K = 10,977 sequences. Via multiplication of a pseudorandom sequence, the dataset is randomized and subsequently encoded with Reed-Solomon codes. Addition of indices (I) and redundancy from the inner encoding step yields N = 16,383 sequences with a length of S = 120 bits, respectively 60 nt.
Fig. 1 Encoding scheme for storing 100 kB of music data in 16,383 DNA sequences. The stored information is structured as 14 packages containing 6 bits (L) with a length of K = 10,977 sequences. Via multiplication of a pseudorandom sequence, the dataset is randomized and subsequently encoded with Reed-Solomon codes. Addition of indices (I) and redundancy from the inner encoding step yields N = 16,383 sequences with a length of S = 120 bits, respectively 60 nt.CC BY · PubMed Central
Fig. 2 Read analysis and comparison to prior work of the photochemically synthesized DNA. a Read length distribution (n = 1). b Normalized GC content c Error probability of reads cut after 60 nt in 5′–3′ direction. d Comparison of errors in previous work. Source data are provided as a Source data file.
Fig. 2 Read analysis and comparison to prior work of the photochemically synthesized DNA. a Read length distribution (n = 1). b Normalized GC content c Error probability of reads cut after 60 nt in 5′–3′ direction. d Comparison of errors in previous work. Source data are provided as a Source data file.CC BY · PubMed Central
Fig. 3 Relative frequency of DNA bases for each position. a Relative frequency as specified by design. Note that the peaks in positions 40–50 nt are due to the index. b Measured relative frequencies (n = 1). c Monte-Carlo simulation with deletion error rate λ = 7 (poisson distributed probability), a substitution error rate λ = 2 and an insertion rate λ = 3. The simulated tailing ratio is C:T 5:1. Source data are provided as a Source data file.
Fig. 3 Relative frequency of DNA bases for each position. a Relative frequency as specified by design. Note that the peaks in positions 40–50 nt are due to the index. b Measured relative frequencies (n = 1). c Monte-Carlo simulation with deletion error rate λ = 7 (poisson distributed probability), a substitution error rate λ = 2 and an insertion rate λ = 3. The simulated tailing ratio is C:T 5:1. Source data are provided as a Source data file.CC BY · PubMed Central
Fig. 4 Spatial error analysis of the photo-directed synthesis machine. a Edit distance of reads to the original sequence mapped to the synthesis position on the photolithographic chip (n = 1). b Number of reads after clustering for each position. 128 × 128 = 16,384-1 sequences were synthesized, resulting in a file size of 99.1 kB. Source data are provided as a Source data file.
Fig. 4 Spatial error analysis of the photo-directed synthesis machine. a Edit distance of reads to the original sequence mapped to the synthesis position on the photolithographic chip (n = 1). b Number of reads after clustering for each position. 128 × 128 = 16,384-1 sequences were synthesized, resulting in a file size of 99.1 kB. Source data are provided as a Source data file.CC BY · PubMed Central
Fig. 5 Clustering and extraction of read DNA sequences. The sequenced DNA fragments are clustered, aligned and possible candidates are extracted in a majority voting step. The alignment is visually enhanced by marking deletion errors with a dash.
Fig. 5 Clustering and extraction of read DNA sequences. The sequenced DNA fragments are clustered, aligned and possible candidates are extracted in a majority voting step. The alignment is visually enhanced by marking deletion errors with a dash.CC BY · PubMed Central
Fig. 6 Read analysis for files of larger size. a, b Read length distribution for file 2 (n = 1) and file 3 (n = 1), respectively. c, d Error probability of reads cut after 60 nt in 5′–3′ direction for file 2 and file 3, respectively. Source data are provided as a Source data file.
Fig. 6 Read analysis for files of larger size. a, b Read length distribution for file 2 (n = 1) and file 3 (n = 1), respectively. c, d Error probability of reads cut after 60 nt in 5′–3′ direction for file 2 and file 3, respectively. Source data are provided as a Source data file.CC BY · PubMed Central
Fig. 7 Comparison of different DNA synthesis platforms and their characteristic traits. Printing technology is primarily used by Twist and Agilent. Electrochemical synthesis is employed by Custom Array. This work uses light-directed synthesis.
Fig. 7 Comparison of different DNA synthesis platforms and their characteristic traits. Printing technology is primarily used by Twist and Agilent. Electrochemical synthesis is employed by Custom Array. This work uses light-directed synthesis.CC BY · PubMed Central

Verification

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Citations

Cites

Show all 30 references

Cited by

Sorted by citations

Items 101–173 of 173

Metadata

Authors
Antkowiak, P. L., Lietard, J., Darestani, M. Z., Somoza, M. M., Stark, W. J., Heckel, R., & Grass, R. N.
Affiliations
  • ETH Zurich (CH)
  • University of Vienna (AT)
  • Rice University (US)
  • Technical University of Munich (DE)
Corresponding author
  • Heckel, Reinhard
  • Grass, Robert N.
Venue
Nature Communications
Published
2020-10-22
Volume
11
Issue
1
Pages
5345-5345
ISSN
2041-1723 (ISSN-L)
OpenAlex
W3093968171
Cited by
173
References
30
Citations by year
2026: 27 · 2025: 44 · 2024: 41 · 2023: 25 · 2022: 18 · 2021: 16 · 2020: 2
FWCI
4.95
Type
journal article
Language
en
Open access
Locations
  • Nature Communications (publishedVersion, cc-by)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • RePEc: Research Papers in Economics (submittedVersion)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
  • mediaTUM (Technical University of Munich) (submittedVersion)
  • mediaTUM – the media and publications repository of the Technical University Munich (Technical University Munich) (submittedVersion)
  • Repository for Publications and Research Data (ETH Zurich) (submittedVersion, cc-by)
  • ETH Zürich Research Collection (cc-by)
Funders
  • Rice University
  • Austrian Science Fund (P27275, FWF P27275, P 30596, P27275 and P30596)
Topics
  • Physical Sciences › Mathematics › Theoretical Computer Science › DNA and Biological Computing
  • Physical Sciences › Mathematics › Theoretical Computer Science › Algorithms and Data Compression
  • Physical Sciences › Computer Science › Computer Networks and Communications › Advanced Data Storage Technologies
Keywords
DNA data storage, information reconstruction, error correction, encoding algorithms, archival storage, solid-phase synthesis
MeSH terms
Algorithms, Base Sequence, DNA, Light, Monte Carlo Method, Gene Library, Information Storage and Retrieval, Sequence Analysis, DNA, Oligonucleotide Array Sequence Analysis, Photochemical Processes, Chemistry Techniques, Synthetic

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review