Distances in mothur (braycurtis, jaccard, anderberg)

Commands in mothur
1

Hi,
I am currently working on my Bachelor Thesis in Austria and I use mothur to calculate distances.
I have one question why do all this distances differ from the original ones?

For example:


Bray Curtis:

mothur(1-2sum(min(SAi,SBi))/sum(SAi)+sum(SBi))

in literature: 2sum(min(SAi,SBi))/sum(SAi)+sum(SBi),


or another example:

anderberg:

mothur(1-SAB/2SA+2SB-3SAB)

and in literature it is SAB/2SA+2SB+SAB)


I hope you could help me with this :idea: ,
Cheers Kerstin Wallner (Technical University of Graz)
2

The ones you are finding are the similarity values - we calculate the distances, which are 1-Similarity

3

Thank you, but i know that this is the reason why the formula is 1-similarity, but that’s not the only difference (please note the underlined info) :

Anderberg:
mothur(1-SAB/2SA+2SB-3SAB)

and in literature it is SAB/2SA+2SB+SAB

Jaccard:
mothur (1-SAB/SA+2SB-SAB)

and in literature it is SAB/SA+SB+SAB

:idea:

4

Sorry for not looking closer at your text. So for us SA and SB are the number of OTUs in communities A and B - these counts include the number of shared OTUs. For your equations, SA and SB are the number of OTUs only found in A and B and are not shared.

5

Have I understood that properly?
(anderberg, jaccard)

mothur:
SA: OTU in A but this OTU could also be in B
SB: OTU in B but this OTU could also be in A
SAB: OTU only in A and B at the same time

Literatur:
SA: OTU only in A
SB: OTU only in B
SAB: OTU only in A and B at the same time

Thank you very much :wink:

6

well it depends on your source in the literature, but yes.

7

Thank you very much, do you know the source for mother’s distances?