While making the second @AsimovPress book, we asked CATALOG (a DNA computing company in Boston) if they could write it in DNA. And they said yes! But the encoding steps, or how the book was converted to nucleotides, wasn't obvious. Now a new paper explains it all in detail. The open-access paper explains how CATALOG wrote "The Complete Works of William Shakespeare—about 1 million words of text—into DNA with raw error rates rivaling conventional media." Now, a direct encoding scheme would be to simply convert each letter of text into its binary form. So you could say that A = 00, T = 01, C = 10, and G = 11, and then you could go through all the words, convert them into binary, and synthesize the DNA accordingly. But that would not be super efficient and would be super expensive to synthesize! So instead, CATALOG started with a plain text file of Shakespeare, where all the punctuation had been removed and all the words were made lowercase. Then, they split this text into individual words, resulting in 982,890 total words or 29,869 unique words. The final file was only 5 MB. Each unique word was next turned into DNA. The trick here is that CATALOG didn't use a direct encoding scheme, but rather started with 81 short pieces of DNA that they had already made ahead of time. You can think of these 81 sequences a bit like Lego bricks. Then, these "bricks" were partitioned into 14 sets; some sets had 4 bricks, others had 5, and and so on. CATALOG took just the first six sets of DNA bricks (containing 6, 5, 5, 5, 5, 5 bricks, respectively) and used that to write the words in Shakespeare. The other sets were used for retrieval, identifying where in the text the words show up, and so on. They picked exactly one brick from each of these six sets and stuck them together in order, thus making a longer DNA molecule called an identifier. There are 18,750 possible identifiers one can make by connecting a single brick from each of the first six sets. I'm simplfying a bit, but each word was basically assigned a set of 3 different identifiers. (So for the word love, maybe identifiers #984, #12,442, and #17,301 would be selected, and they would synthesize and combine those 3 DNA molecules.) Every unique word in Shakespeare got its own unique pattern of three identifiers, and that pattern is the code for the word. So finally, they went through the Shakespeare text word-by-word, looked up the three identifiers for each word, and told their custom-built DNA printer — a machine called “Shannon” — to make them. Shannon works fast because it just mixes the bricks together into a spot and lets chemistry glue them together automatically. The DNA printing took minutes, and processing the DNA afterwards took a few hours. They checked their work by sequencing the DNA and comparing it to the original text. Only about 2% of the identifiers were wrong before error correction. Read the paper: dl.acm.org/doi/pdf/10.1145/3… Get the book ($30): press.asimov.com/books

Asimov Press. Book Unboxing. Encoded in English and DNA. Designed by Everything Studio | Brooklyn, NY Printed by Shapco | Minneapolis, MN Published by Asimov Press | San Francisco, CA CATGATGACAGCTAGCGTCTAAGCCATTGTTGTTCGACG

More photos of the @AsimovPress book. Our book printer (based in Minneapolis) had to custom-make several parts for each order, including a display stand to hold each DNA capsule. The DNA was made by @catalogtechinc, sealed by Imagene in France, and sequenced by @plasmidsaurus.