User Experience Project Structure Design
In this section, we describe the user experience project structure design of the PHYLOViZ Web Platform. We describe the vision of what the user will be able to do within the app's project structure and what features are to be available.
- Creating a Project
- Uploading Files
- Creating a Dataset
- Computing and Visualizing
- Multiple Views and Algorithms
- What's new - Main differences from PHYLOViZ ONLINE and PHYLOViZ (desktop version)
Firstly, the user must create a project, giving it a name. Withing the project, they can upload files separately and create datasets.
An empty project file structure is shown to the user, containing no uploaded files and no datasets.
The user can opt to first upload the files, and only after that create a dataset selecting these pre-uploaded files. The files will remain in the project, not associated directly with a single dataset. This allows for these files to be used in multiple datasets.
The files fall under two categories: Typing Data and Isolate Data, and on the moment of the upload, the correct type of the file must be chosen to be associated with the file.
To create a dataset, which will hold all the resources of a certain study, a typing data must be selected to be associated with the dataset. The dataset has one and only one typing data. An isolate data file can optionally also be selected (again, only one).
This selection of data can be through the pre-uploaded files (shown here), or through uploading directly in the dataset creation.
Notice below that the isolate data and typing data are linked.
Clicking on the dataset will show a list of options:
- Launch Tree - Launches a tree view, choosing an already existent tree or computing a new one and showing the tree view at the end.
- Compute Distance Matrix - Computes a distance matrix.
- Compute Tree - Computes a tree, choosing from an existent distance matrix or computing a new one.
- View Typing Data - View the typing data as a table.
- View Isolate Data - View the isolate data as a table.
When choosing the compute distance matrix, compute tree or launch tree options, the project file structure will update as shown in the below images.
Just like the dataset, clicking in any of the resources will show a list of options:
- View - View the distance matrix. Multiple view options may be available (table view, heatmap view).
- Compute tree - Computes a tree, which is the next step.
- Launch tree - Go through the rest of the computational steps (calculate inference tree, calculate coordinates) and show the tree at the end.
- Download - Download the distance matrix as a file.
- View - View the tree. Multiple view options may be available.
- Launch tree - Compute and show a tree view.
- Download - Download the tree as a file.
- View - View the tree view (in the specified layout, with filters, etc).
Clicking a file (typing or isolate data) shows options too:
- View - View the data in a table.
- Download - Download the file.
Whenever an option that computes a tree view is selected, a new Tree View is created, possibilitating multiple views even if the same tree was used. Notice in the image below that there is only one existing tree, but multiple views.
Each resource is identified with a source specifying how it was loaded into the dataset; might be by computation/algorithm or by direct upload e.g. newick file. Also, provided the option to choose what algorithm is to be used for each step, this may result in multiple algorithms being used, and being appropriately organized in the project structure.
In the online version, it's already possible to view the distance matrix as a heatmap, but the limitation is that it's necessary to compute all the steps and launch the tree, and only after that the possibility of generating a heatmap based on selected nodes is given. In this new version, the aim is to provide an easy way to compute and visualize only what the user deems as necessary.
In the desktop version, it's already possible to to generate multiple tree views for a same project. However, these views encompass all the intermediate steps. In this new version, it is important for the intermediate steps to be shown and carefully organized, providing a way for multiple algorithms to be used in different combinations.