tag:blogger.com,1999:blog-4503292949532760618.post7660995589753446315..comments2024-03-16T18:42:21.178-07:00Comments on DSHR's Blog: Nature's DNA storage clickbaitDavid.http://www.blogger.com/profile/14498131502038331594noreply@blogger.comBlogger2125tag:blogger.com,1999:blog-4503292949532760618.post-4623051363194871072017-03-05T13:44:30.956-08:002017-03-05T13:44:30.956-08:00Last week Science published DNA Fountain enables a...Last week Science published <a href="https://doi.org/10.1126/science.aaj2038" rel="nofollow"><i>DNA Fountain enables a robust and efficient storage architecture</i></a> by Yaniv Erlich and Dina Zielinski from Columbia. They describe an improved method for encoding data in DNA that, at 1.83 bits/nucleotide, gets much closer to the Shannon limit of 2 than previous attempts. Their experiment stored 2.2MB at about $3500/MB write cost.<br /><br />The authors admit that write cost is a problem, but:<br /><br />"we envision that the cost issue of DNA storage could be addressed by two complementary approaches, the first of which is continuous improvements to the DNA synthesis chemistry, which have been estimated to exponentially reduce the cost by one to two orders of magnitude per decade (4)."<br /><br />Reference 4 is the 2013 EMBL paper, whose cost claims <a href="http://blog.dshr.org/2013/01/dna-as-storage-medium.html" rel="nofollow">I debunked at the time</a>. Rob Carlson's <a href="http://www.synthesis.cc/synthesis/2016/03/on_dna_and_transistors" rel="nofollow">blog post</a> shows that 1-2 orders of magnitude/decade is a vast over-estimate, even ignoring the fact that magnetic media is getting cheaper faster.David.https://www.blogger.com/profile/14498131502038331594noreply@blogger.comtag:blogger.com,1999:blog-4503292949532760618.post-73889019697057076182016-09-21T12:33:19.682-07:002016-09-21T12:33:19.682-07:00There were some interesting discussions of DNA sto...There were some interesting discussions of DNA storage at the <a href="http://www.digitalpreservation.gov/meetings/storage16.html" rel="nofollow">Library of Congress Storage Architecture workshop</a> and there was general agreement on two points:<br /><br />- DNA synthesis needs to get something like 6-9 orders of magnitude cheaper for DNA to be a cost-effective archival medium.<br /><br />- Although there is scope for improvement, especially as synthesis for storage needs lower accuracy than for current markets, it isn't likely that current synthesis techniques can ever be improved that much. Radically different, more "biological" techniques would be needed.David.https://www.blogger.com/profile/14498131502038331594noreply@blogger.com