Commenting code

[mvisani]

Hey, I have a quick question. I am currently trying to comment the code of the different scripts in https://github.com/anticipated-lotus/elotus-trees. I am not sure to understand the outputs of prepare_pseudoabsence_tables.R.

• What do "ref" and "no_ref" or "children" or "no_children" mean ?
• What are the differences in the outputs ?
• Are the numbers in the outputs the number of references, occurrences ?

[Adafede]

Hey,

First of all, very sorry for the current state of the code and documentation!
Everything was done very quickly and dirtily so do not hesitate to ask/change anything.

Regarding your question, "ref/noref, children/nochildren" mean if you take into account multiple reference (and children) or not.

So, for example: 2 articles report amarogentin in Gentiana lutea, 3 (of which 1 is the same) in Gentiana acaulis.

Then, for the pair amarogentin in Gentiana:
ref_child: 5
ref_nochild: 4
noref_child: 2
noref_nochild: 1

Again, this might make sense or not at all depending on what you will be doing afterwards. Everything is just to get somehow a sense of the data in an exploratory manner for now. 😊

[mvisani]

Hey ! Thanks for the reply !

So if I understood this correctly we would have the following ?

• ref_child : 5 → Gives the number of occurences were we find amarogentin related with a Gentiana species.
• ref_nochild: 4 → Gives how many different papers there are that talk about amarogentin in the species Gentiana
• noref_child : 2 → Gives how many children of that species were found with at least one reference (in that case : Gentiana
  lutea and Gentiana acaulis so $2$ )
• noref_nochild: 1 → Will only be either 0 or 1. Gives if it there is at least one reference that talks about that molecule
  in the species Gentiana

[Adafede]

Yes, I just realized the tables are grouped also chemically. So child also refers to the "chemical children", so exactly same principle as for Gentiana and G. lutea but this for Terpenoids > Monoterpenoids > Secoiridoids (> Amarogentin)... Sorry everything is quite old but do not hesitate if still unclear!

[mvisani]

Makes sense, thanks !
I've made some changes in the code in order to also accept 3D structure of molecules. Please tell me if you run into some problems ;) (normally it should work)

[Adafede]

You directly started making changes, wow, very cool! 😊

I did setup some tests for changes to be checked automatically but as the repo is private, we only have 60min/month for free (see https://github.com/anticipated-lotus/elotus-trees/actions/runs/4156288192).

Depending on how the project evolve we could change this or make the repo public to overcome the limit? @oolonek

Anyway, very cool! I am just looking from times to times but if you need something more concrete from me or would like to discuss something, do not hestitate!

[oolonek]

Right ! And not just regarding gh actions limits. Done now.

[Adafede]

Very cool, thank you!

I did some additional cleanup to avoid confusion as much as possible. Just ping whenever you need! 😊

[Adafede]

Hey @mvisani, I did update things to run again, I guess letting at the class level and not InChIKey is better to avoid a super large matrix which can cause memory issues (at least for now)