On GISAID they classified the corona using 4 clades(S, G, V, Other).
I would like to know exactly how these genomes were classified for my research.
So how do you classify a coronavirus genome as belonging to one of the clades.
enter image description here


2 Answers 2


*The question is a duplicate of this one.

There are many techniques, all encompassed in what is known as molecular phylogenetics. To get 'clades' or any other groups, you need the sequences of your organism (a virus, in this case, each sample will have its own sequence), and then you need to align these sequences (to match the regions that can be compared, remove 'blanks', 'duplicates', etc). The alignments will differ in some informative positions, and these differences will give you an estimate of how close are some sequences to others, forming groups (or 'clades' in technical terms). So, how do you actually do this? I don't think Stack Exchange is the proper forum for this if you are doing scientific research (I'd recommend you talk to your advisor or a colleague). But if your research is just for curiosity and you need some basic place to start, I'd recommend you to check out MEGA, and this paper. There they explain, step by step, how would you go about this.

And where do you get the sequences? Well, if you are looking at a specific paper they usually give you the 'accession number' of the sequence(s) they are reporting, and if you just want to find your own sequences for your own study, you can search the GeneBank.

  • $\begingroup$ Thanks for the answer. I already have the genomes and I already have done many alignments in the past so this is not new for me. What I need to know is how to exactly classify a coronavirus genome as belonging to one of the clades mentioned in the question. I asked a similar question before and got a very appropriate answer. This is the kind of answer i'm looking for(In the link). biology.stackexchange.com/a/90896/57497 $\endgroup$
    – yuval
    Commented Apr 5, 2020 at 12:05
  • 2
    $\begingroup$ If this is for serious research I'm sure you have people who could really help you in your university/department, and not a random stranger in Stack Exchange. However, I noted that the answer you are looking for is already in the top left corner of your image. The criterion the authors of the study used is there in plain sight. Good luck! $\endgroup$ Commented Apr 5, 2020 at 12:11

The clades have an origin point from which they can see changes in the virus structure to define the clades themselves. The origin point, or reference is the established SARS-CoV-2 reference genome, NC_045512. The clades are mainly named after the mutations that amino acids and proteins have had.

Source: https://www.who.int/bulletin/volumes/98/7/20-253591/en/

They look for different types of mutations and what variants in the SARS-CoV-2 lead to these. Sequences are submitted from across the globe and uploaded to GISAID and analyzed. For example, the S and L variants. There's been a change in the amino acid leucine to serine. So the clade encompasses both but the change to serine (S) defines the clade.

For ORF8-L84S (as seen in picture), the ORF8 gene underwent a mutation of the amino acid L at the 84th position of the gene’s protein to the amino acid S. In other clades, such as G changes aspartic acid at position 614 to glycine. Clade V is labelled NS3-G251V, where the NS3 gene’s glycine at position 251 changed to valine.

So for the naming in general it goes like this:

Clade S - ORF8-L84S

ORF8 (gene undergoing mutation)

L84 (L indicates amino acid leucine at position 84)

S (last letter indicates what it changed into, and defines the clade)

Here's a picture of the SARS-CoV-2 encoding sequence structure and location of common mutations: enter image description here Source: https://www.biorxiv.org/content/10.1101/2020.05.03.066266v1.full

The mutations define the clades. You collect the sequences and run a maximum likelihood (ML) phylogenetic. This will show you how the sequences are clustered compared to your reference genome. You can then do a lineage classification software (Pangolin COVID-19 Lineage Assigner).

So for classifying the clades they align the struces with reference genome NC_045512. They look for changes or mutations in the virus and separate them into clades. This might also be done by functional and non-functional mutations but I am not sure if that is included into GISAID.


You must log in to answer this question.

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