My goal is to do a de novo assembly of a jellyfish. At the moment I'm busy checking papers that did de novo assemblies but I also got some jellyfish papers Then I stumbled accross this:

Contig NG50 is defined as the length at which half of the predicted genome size is contained in contigs longer than this length.

I can’t understand this and googling returned papers of similar complexity e.g.:

NG50 refers to the length of the smallest contig added to cover 50% of all nt estimated in the genome; (25)...

Can someone provide a simpler explanation?

  • 3
    $\begingroup$ Please do not direct users of this site elsewhere for the meaning of specialist abbreviations central to your question. Questions here should be self-contained, and if you request clear answers, I suggest you do as you would be done by. $\endgroup$
    – David
    Commented Jul 3, 2021 at 18:15
  • $\begingroup$ The Wiki entry "N50, L50, and related statistics" would be a good research resource. en.wikipedia.org/wiki/N50,_L50,_and_related_statistics $\endgroup$
    – Armand
    Commented Jul 3, 2021 at 20:03
  • 1
    $\begingroup$ This stack is for professional/student biologists, not 5 year olds, so I'm unsure why you seem so insistent on on dumbing down the title. $\endgroup$
    – user438383
    Commented Jul 5, 2021 at 20:04
  • 1
    $\begingroup$ @user438383 Because my professors always told me only truly understand something if you can explain it to your 5 year old brother :) And also because I thought this is a positive and supportive place where people are encouraged to ask questions that other people are encouraged to answer (collecting votes I assume). Have you ever read a paper that was not in your native language? It can get very tough, but I hope that I understand a little bit more every day, with the kind help of smart and motivated people like you and other members of this community, but maybe I'm just a naive dreamer :) $\endgroup$
    – ilam engl
    Commented Jul 5, 2021 at 20:21
  • 2
    $\begingroup$ I have edited your question to provide a partial model of good practice. (The first sentence is too chatty.) If your aim is to communicate the results of your (genome?) assembly you would do well to study it. $\endgroup$
    – David
    Commented Jul 5, 2021 at 21:43

1 Answer 1


If you arrange your contigs from longest to shortest, what is the shortest contig in the smallest subset of contigs that covers at least half the expected genome length?

An example with arbitrary units...

You have a set of contigs of lengths 30, 20, 20, 10, 5, 5, 1, 1, 1, and many contigs < 1. Your expected genome length is 150. The smallest subset of contigs that covers at least half the genome length is [30, 20, 20, 10]. The smallest contig in that set has length 10, so NG50 = 10.

Note that the sum of the lengths of the contig subset is 80, not 75. But if you take away the smallest contig in that subset (10), then the sum is 70, which is less than half the genome length. So the sum of the lengths of the contig subset need not be exactly half the expected genome size, but at least half.


You must log in to answer this question.

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