Gupta and Mohan

cleaved. All resultant fragments in the library are not only suitable for separate

regions of the active site, but also maintain their original binding modes and

conformations of the initial known inhibitors within the protein target active site.

New compounds generation

We used LigBuilder software to generate the new compounds. First, it analyzes the

AChE binding pocket using the POCKET module of LigBuilder software (Wang et

al, 2000). The pocket algorithm generates a pocket and the grid file which is used

as an input for running GROW or LINK module to build ligand from the fragment.

It also generates key interaction sites based on the key interacting atoms of the

important active site residues.Along with this file a structure based pharmacophore

is also generated. For pharmacophore generation the minimum distance between

the features was kept at 3.5 Å, so as to avoid any steric and electrostatic clashes

between protein residues and designed ligands. Maximum number of features in

the pharmacophore was kept as eight, as more features will not be significant.

After generating a pharmacophore map and the key site map by pocket module,

next step is to build lead compounds by using these chemical features and key site

information of AChE active site. To calculate the binding affinity, this program

uses the SCORE algorithm. SCORE is an empirical procedure developed to

estimate the binding free energy of a compound to its receptor protein when the

3D structure of the complex is known. PROCESS module ranks the compounds

generated by the GROW module based on the SCORE function in the LigBuilder,

which extract the desired compounds, and convert them to viewable mol2 files.

LigBuilder has couple of built in chemical criteria to avoid “bad structures”, such

as -O-O-, -N-O-, -N-N-, and etc. Compounds meeting all these criteria will be

considered for output. The new compounds were further evaluated by theADMET

filters.

ADMET filtering

We have computed the Lipinski’s rule-of five and other pharmacokinetic properties

of the compounds, which includes (Egan et al , 2004): distribution coefficient

(Log D), computed aqueous solubility (Log S), polar surface area (PSA), percent

human oral absorption, BBB penetration, and CNS activity using ADME module

of Discovery studio DS2.5 software (Accelrys: 2.5, CA, 2006). The rule-of-five

suggested that a chemical compound could be an orally active drug for humans.

Toxicity profiles of the compounds were assessed using DEREK software (DEREK

10.0, U.K., 2010) . DEREK is a knowledge-based expert system for the qualitative

prediction of toxicity. DEREK makes its predictions based on a series of rules and

each rule describes the relationship between a structural feature and its associated

toxicity (DEREK 10.0, U.K., 2010) . Further, docking analysis was performed on

116

Int. J. Bioinformatics and Biological Sci.: v.2 n.1&2, p. 111-126. March & June 2014