The main reason to make the user interface is so that users can use models that were pretrained. For cheminformaticians the user interface and underlaying package can also be used to share additional models that they have trained themselves.
The way that models are displayed could be improved
- currently not in alphabetical order
- the context of the AOP that they are a part of is lost
- if more case studies are added the tabs that we are currently using won't work anymore because it would become to cluttered
I forsee two extremes in terms of model sharing
- not related to AOP at all, you just have the name of the endpoint that the user want to predict (this can be a MIE but doesn't have to be)
- very much related to the AOP, show the whole adverse outcome pathway network including the MIEs (in a visual manner). have the user select which MIE they want to predict. See image below for example
Next steps;
- at least put in alphabetical order
- decide on how dependent we want to make the model sharing on AOPs
The main reason to make the user interface is so that users can use models that were pretrained. For cheminformaticians the user interface and underlaying package can also be used to share additional models that they have trained themselves.
The way that models are displayed could be improved
I forsee two extremes in terms of model sharing
Next steps;