Home
This Title All WIREs
WIREs RSS Feed
How to cite this WIREs title:
WIREs Comput Mol Sci
Impact Factor: 25.113

Enabling future drug discovery by de novo design

Full article on Wiley Online Library:   HTML PDF

Can't access this content? Tell your librarian.

Abstract Computer‐assisted drug design is evolving as a source of innovation for drug discovery. In particular, de novo design approaches are being increasingly applied to find novel drug‐like compounds, molecular scaffolds, and bioisosteric replacements for established or unwanted fragments. Although some of the early software tools had a certain tendency to generate compounds of limited chemical attraction, modern de novo design algorithms put a strong emphasis on the synthesizability and drug‐likeness of machine‐generated compounds. We give an overview of the various methodologies for virtual compound construction, evaluation, and optimization in machina, and how they can support medicinal chemistry projects in the early phase of drug discovery. © 2011 John Wiley & Sons, Ltd. WIREs Comput Mol Sci 2011 1 742–759 DOI: 10.1002/wcms.49 This article is categorized under: Computer and Information Science > Chemoinformatics

Computational counterparts for conventional drug discovery methods.

[ Normal View | Magnified View ]

Example of a potent inhibitor of HIV‐RT designed with the help of the software BOMB based on a fixed compound core structure.126

[ Normal View | Magnified View ]

Compound 1 designed by TOPAS and its close structural analog 2 were synthesized and potently block the voltage‐gated human K+‐channel Kv1.5.54

[ Normal View | Magnified View ]

The two most potent inhibitors designed by the Skelgen software for human estrogen receptor α.123,124

[ Normal View | Magnified View ]

The software LUDI was applied to designing an inhibitor of the HIV‐RT.122 Starting with the structure of the binding site, LUDI first placed fragments into subregions of the pocket and linked them by a phenyl moiety in a second step. The resulting structure was finally decorated with an amide side chain by LUDI. Subsequent manual optimization exchanged the pyrrole ring with an imidazole to simplify the chemical synthesis. A series of compounds based on this scaffold has been synthesized and led to different active molecules for several enzymatic activities of the target (the given IC50 value has been determined with respect to DNA polymerase activity of HIV‐RT).

[ Normal View | Magnified View ]

A set of solutions (gray circles) for a two‐dimensional optimization problem. The figure next to each circle denotes the number of dominating solutions. For one solution, exemplary determination of the number of dominating solutions is presented (dotted lines). Three solutions are dominating because they are better in all objectives. Nondominated solutions form the Pareto front (dashed line, adapted from Ref 10).

[ Normal View | Magnified View ]

The software DOGS80 generates new compounds in a reaction‐driven, fragment‐based fashion so that the designs mimic a given reference compound (Raloxifene, a modulator of the human estrogen receptor106). Together with the designed molecules, DOGS also delivers a potential synthesis route. Where appropriate, general names of the reactions are given next to reaction arrows. This reaction‐based approach has a good chance to result in synthetically feasible compounds.

[ Normal View | Magnified View ]

Eleven cleavable bond types defined by RECAP102.

[ Normal View | Magnified View ]

Markov chain model for compound assembly. A Markov chain is a sequential graph traversal: each time a node is visited (indicated by a dashed line) the corresponding building block is added to the growing molecule (bottom). Edge labels (weights) correspond to the probability to walk along an edge in order to get to the next node. Weights are determined by connection statistics of fragments observed in a training set of molecules. Please note that this graph represents a simplification Typically, weights will not be symmetric, which leads to a bidirected instead of an undirected graph as shown.

[ Normal View | Magnified View ]

Fragment linking and growing in receptor‐based (‘in situ’) de novo design. (a) The growing strategy (a) starts from a single preferred fragment and sequentially adds new building blocks to yield the final product. Here, a three‐step growing process is shown. (b) For fragment linking ‘interaction hotspots’ in the target cavity are saturated with preferred building blocks, which are connected by suitable linker fragments. Both approaches can be combined during de novo design. For this illustration PDB structure 3lg499 was used. Note that the compound shown was not actually designed de novo.

[ Normal View | Magnified View ]

Generation of compound chimera by fragment swapping. Original ligands A and B (top) are aligned in a first step (center). The software BREED60 identifies strategic bonds (highlighted in red) and swaps fragments (A1, A2, B1, B2) at this position to yield hybrid structures A′ and B′ (bottom).

[ Normal View | Magnified View ]

Related Articles

Drug Discovery: An Interdisciplinary View

Browse by Topic

Computer and Information Science > Chemoinformatics

Access to this WIREs title is by subscription only.

Recommend to Your
Librarian Now!

The latest WIREs articles in your inbox

Sign Up for Article Alerts