Using two different approaches, we created four versions of a draft mouse model.
In this approach, we create a single combined principal component space for all the classes.
For the mapping between the ontologies of our approach, we create a procedure that corresponds to the previous definition.
In this approach, we create a population of (initially random) hash functions, and each cycle they are evaluated for "fitness" based on their expected entropy and computation cost (method described in Materials S1).
To test our approach, we created, simulated and adapted several 3D models.
Using this approach, we created a total of 16 major design prototypes, with multiple iterations in-between.
Using the ligand-based approach, we created pharmacophores selective for agonist ligands and pharmacophores selective for antagonist ligands for each NRLiSt BDB datasets that yielded high performances.
For each sampling approach, we created a median estimated ACD map across all of the sampling grids (systematic sampling, 5 100 km) and plot sets (stratified random, {1, 5, 10, 50, 100} plots).
Using this approach, we created a list of 5,210 in-domain key phrases that are later used in the block key phrase identification and measuring similarity between two documents.
In the presented approach, we created core-shell-corona micelles with comparable size and shape, but a variable density of d-mannosylated chains sticking out from the surface, and investigated their interaction with the lectin concanavalin A (Con A).
