10
$\begingroup$

The human genome is about 770 MB, the C. elegans genome is about 100 MB, the yeast S. cerevisiae is about 12 MB. Different other genomes have been sequenced: how many GB of genomic DNA we have now?

Let say we would like to make a Noah hard-disk ark: how much space would it take to represent the genomes of all known species on earth? There is a way to provide an estimate?

I'm also interested in the total biodiversity: for instance, if two species each have 1 GB genomes and have half of their DNA in common this would count as 1.5 GB.

$\endgroup$
11
  • 4
    $\begingroup$ I think this is an unanswerable question - there are a handful of species whose genomes have been sequenced, and many millions on the earth that remain completely uncharacterised! $\endgroup$
    – Luke
    May 30, 2012 at 8:42
  • 7
    $\begingroup$ The human genome has highly repetitive (“low complexity”) regions though, which means that it can be compressed much further. This is the real measure of biodiversity “size” since it measures the actual information content. And adding genome sizes is much more complex than you make it out, your 1.5 GB approximation isn’t very precise. $\endgroup$ May 30, 2012 at 10:43
  • 1
    $\begingroup$ @KonradRudolph You have to choose between the most reductive representation of the genome and the fullest one. I'm sure we'll have records of hundreds of millions of variants in the human genome alone by the end of the next decade. There are proposals to move away from a single reference build for the human genome. Not only storing the sequence itself, but annotation describing how the variants are used already blows the doors off of 770 MB. $\endgroup$
    – shigeta
    May 30, 2012 at 16:20
  • $\begingroup$ @shigeta I don’t think this is relevant here. What Craig wants to measure is the combined information content of all genomes, and the information content is by definition the entropy which, again by definition, corresponds to the size of the data in its tightest (theoretically) possible representation. $\endgroup$ May 30, 2012 at 16:23
  • 2
    $\begingroup$ If you want to represent all genetic information on Earth, you need to get all individuals, not all species. There is significant variation between two humans or two birch trees, so you can't just take a single sample per species and say 'got it!'. $\endgroup$
    – Peteris
    Dec 22, 2014 at 7:49

2 Answers 2

14
$\begingroup$

If you simply take one order of insects, Coleoptera, there are just under 400,000 described species with estimates from 850,000 to 4,000,000 species total in just this order. The number of primates is under 1,000. If your assumption of say 10MB for all other primates would be accurate, just adding in the low end estimate of 850,000 at 10MB per 1000 we are quickly at 8,500GB which seems to be a factorial out of the GB range.

So, we have a broad estimate of non-bacterial of plants, animals etc. at say 8,700,000.

Jason Gans found in a 1 gram of soil survey approximatly 1,000,000 bacterial species.

SO the total accounting for species number is totally impossible to estimate for anything at this time, let alone the genome.

Even for something as "common" as a giraffe, there are up to 9 sub-species with genome differences within each subspecies.

So, once we get them all decribed, we can then work on the genome sequence for each and get you some answers!

$\endgroup$
1
  • $\begingroup$ Recent genetic work on the Giraffe now indicates multiple species of what were once thought to be sub-species, some of which are now viewed as endangered of extinction due to the low population numbers in some. $\endgroup$ Jan 18, 2017 at 17:16
2
$\begingroup$

According to this paper there is ~ $5.3 \times 10^{34}$ base pairs of DNA on earth. One base pair contains approximately 2 bits worth of information (4 possibilities). So that would be ~ $8.48 \times 10^{35}$ bytes of information.

or more than the total information in all digital and analog technological storage media on earth combined.

$\endgroup$
2
  • $\begingroup$ I was mostly wondering about thy total information content in all dna on earth. This number is probably much higher than the total information content as many of those strands will be duplicates, but I upvoted your answer as this still provides a good upper bound and possibly some good places to start with other calculations. Thanks! $\endgroup$
    – Craig
    Sep 13, 2018 at 19:32
  • 1
    $\begingroup$ duplicate information is still information, most of the information in books or on digital media is duplicate as well. Any estimate you compare it to will contain a lot of duplicate information. Alos the original paper lists the number in megabases, I converted for convenience. $\endgroup$
    – John
    Sep 14, 2018 at 2:31

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .