INVDOCK performs docking of low-weight ligands into cavities of each target

using a computer-automated search for potential protein and nucleic acid targets. It

ranks the ligands based on molecular mechanic interaction energy and competitive

binding analysis (Chen and Zhi 2001). TarFisDock is another inverse docking tool

(Li et al. 2006) and ranks the ligand-protein interaction in terms of binding energy

(Shoichet and Kuntz 1993). idTarget optimizes search space by dividing the poten-

tial target into small boxes based on the size of the ligand followed by identification

of binding sites using an optimized MEDock algorithm (Wang et al. 2012). Conven-

tional docking programs such as AutoDock Vina and Glide software (McMartin and

Bohacek 1997; Morris et al. 1998) have also been adapted to incorporate the feature

of reverse docking (Rognan 2010).

2.2.5

Multi-Target Drug Design (MTDD)

This approach is promising for neurological disorders and cancers that are complex

multifactorial diseases. Better therapeutic efficacy and safety is known to be

achieved by designing individual new chemical entities that can simultaneously

target different points of a given pathogenic cascade. MTDs have been shown to

have a higher synergistic effect as compared to a combination of drugs (Bottegoni

et al. 2012). They are developed using either of the two available strategies: a

fragment-based approach, involving the combination of pharmacophores from selec-

tive, single-target ligands, and a single, multitasking computational model, involving

screening of compound collections to identify compounds with a suitable combina-

tion of activities by simultaneous application of multiple computational models. A

hybrid molecule with a dual mode of action that has been designed is donecopride,

which is a novel drug candidate for Alzheimer’s disease that has been shown to

exhibit dual binding site inhibitory effects (Lecoutey et al. 2014).

2.2.6

Multi-Target Virtual Ligand Screening (VLS)

Rational drug design project to identify multi-target hits can begin following the

identification and validation of a suitable combination of targets. High-throughput

screening (HTS) can be successfully used to identify initial hits but is time-

consuming and expensive for even one target and much more cumbersome when

multiple targets are to be considered simultaneously. Therefore, as an efficient and

faster alternative to HTS, virtual ligand screening (VLS) is being used for processing

large libraries of compounds (Abagyan and Totrov 2001). In VLS, every molecule in

the library is tested against an ideal model of activity, and they are ranked by

assigning each compound a predicted activity score. Only the top-ranking fractions

are analysed using further testing (Jenwitheesuk et al. 2008). VLS applied to multi-

targets thereby helps to identify hybrid molecules that can simultaneously bind to the

selected targets. Wei and co-workers successfully identified novel anti-inflammatory

candidates displaying activity against phospholipase A2 (PLA2) and human

2

Polypharmacology: New Paradigms in Drug Development

21