Table 2.

Input variables (containing descriptors) and output variables (containing binding scores activities) of CPANN models.

Input variables	Two descriptors: polarizability volume in cubic angstroms (QPpolrz) (1) and topological diameter (TD) characterized the size of molecules (2);
	Monte Carlo descriptors or so-called optimal descriptor (DCW) (1);
	Twenty five drug like descriptors: H-acceptors (1), H-donors (2), total surface area (3), relative PSA (4), polar surface area (5), drug-likeness (6), Mol. Weight (7), cLogP (8), cLogS (9), electronegative atoms (10), stereo centers (11), rotatable bonds (12), rings closures (13), small rings (14), aromatic rings (15), aromatic atoms (16), sp3-atoms (17), symmetric atoms (18), amides (19), amines (20), aromatic nitrogen (21), basic nitrogen (22), acidic oxygens (23), non-H atoms (24), non-C/H atoms (25).
Output variables	Average Binding Scores (1)
	Binding Scores for five proteins related to diabetic disease:1BMQ (1)- Enzyme/Hydrolase;1FM6 (2)- Receptor/Transcription;1GPB (3)- Enzyme/Glycogen phosphorylase;1H5U (4)- Enzyme/Glycogen metabolism;1US0 (5)- Enzyme/ Oxidoreductase.