Add DisProt features

[CLPF]

• Fix annotation fetching not trigerring render
• Add disorder features
• Add AlphaFold track
• Add transition and functional information inside features

[colin-combe]

Hi,
thanks for your contribution,
I'm not sure about ef2ab18

it would cause re-rendering annotations and the key after every protein's individual annotation load rather than when they have all loaded?

Colin

[EliotRagueneau]

Hi @colin-combe , that commit actually comes from me.

originally as you incremented the counter once per annotation group, you would reach a total of n annotation group * p proteins, e.g. 10 proteins and 2 annotations (superFam and Unitprot) would lead to a total counter of 20, but you would only trigger the rendering midway through, let's say when all superfam have been processed, or 5 superfam and 5 uniprot.

We noticed this issue while adding DisProt because they were never being rendered with the current system.

With the proposed commit, once one "track" has fully loaded all its proteins, we render it.

We could also wait for all the annotations from all the resources to be loaded before rendering them, but then you risk having uniprot features not being displayed if superFam is not working for instance, or just taking a long time to resolve.

[colin-combe]

@EliotRagueneau - right, i see now it was broken.

With the proposed commit, once one "track" has fully loaded all its proteins, we render it.

Is that the behaviour now? Or is the protsAnnotated++ counter now redundant and it renders every time a request returns (after all requests have been sent)?

cheers,
Col

[EliotRagueneau]

Is that the behaviour now? Or is the protsAnnotated++ counter now redundant and it renders every time a request returns (after all requests have been sent)?

Well we only trigger the callback once the counter reach the number of proteins, on the last callback after annotations have been created, so yeah, basically when the last protein have received back it's annotation from Uniprot, then we render everything. When the last protein received its annotations from SupFam, we render. The counter is actually important to keep track on whether we received the last protein or not 😄

[colin-combe]

The counter is actually important to keep track on whether we received the last protein or not 😄

I don't think so; there's nothing keeping track of whether responses are received.

The protsAnnotated counter is incremented on ln 17 of annotation-utils, in the synchronous block before the calls to the functions that will send the asynchronous requests (lines 21, 24 and 27).

So protsAnnotated is currently a count of proteins that requests have, or soon will be, made for and not of responses to requests that have been received.

Right?

best wishes,
C

[EliotRagueneau]

Hi Colin,

The counter is being compared after the annotations have been added, so once the annotations have been received from the external resources.
That being said, it's true that for the last one, say we have 20 proteins, the 19th could reach the callback before the 20th have been processed, but after the counter has increased to 20, and therefore trigger the callback a bit too early. I'll try to find a better solution, thanks for pointing that out!

[colin-combe]

Hi Eliot,

the code was equivalent to:
let protsAnnotated = 0; const molCount = 3; for (let i = 0; i < molCount; i++) { protsAnnotated++; Promise.resolve().then(() => console.log(protsAnnotated === molCount)); }
which will print true, true, true

protsAnnotated will equal molCount before any of the async functions are called, if you want to see that then add console logs to see the vaules of protsAnnotated and molCount when the condition is checked