Widgets are the toolbox of the data visualization space. They give the user freedom to set up their own visualization layout, by selecting the tools for data view, analysis and visualization of their own choosing.
Widget menu
To see and select widgets, click on the Widgets button in the top bar of the data visualization space. The widget drawer opens and lets you see which widget are available for use.
Widget customization
To customize widgets, use the Settings button in the top bar of each individual widget. This allows you to change the scores that are displayed, the parameters of the graphs, etc.
List of available widgets
Overview
Help
Drag this on the layout to learn how to resize and move your widgets around.
Generator details
See the details of your generation set-up (starting fragments, substructure constraints…) at a glance.
Data View
Table & Cards
The main tools for data visualization. Users can easily switch between Cards and Table view. More information is given in the Molecule View article.
UMAP
Visualize and explore your chemical space in 2D. Makya applies UMAP dimensionality reduction to a pairwise molecular distance matrix, projecting similar molecules closer together in a 2D space. Clusters are then identified using the DBSCAN algorithm. Use the lasso to select molecule subsets.
Design
Sketcher
Draw any molecule and compute its physico-chemical properties with RDKit in real time. Add the molecule to the current dataset, where it can be rescored or used as a reference molecule.
Filters & Analysis
Parallel Coordinates
Another way to filter your compounds! Through the gear icon for setting, you can customize what to include in a coordinate.
Plots
2D Chart
Visualize two variables on the X and Y axes. Use color to map a third variable and shape for a fourth one.
3D Chart
Visualise three variables on the X, Y, and Z axes! You can add color to represent an extra dimension. Rotate, zoom, and pan around to explore the spatial patterns, clusters, or outliers in more detail.
Distribution
Visualize how a single variable is distributed. Choose a dimension and set the number of buckets to adjust the resolution of the histogram.