A platform for the Design of Synthetically Accessible
Molecules with Desirable Properties in Medicine
and Material Sciences
The Synopsis Platform
Synopsis is a de novo design platform that generates synthetically accessible
molecules with desirable properties.
The properties are calculated by a scoring function that is independent from the molecule
generation. The scoring function can be based upon any proprietary or commercially available software.
As such, Synopsis can be deployed in structure-based and ligand-based drug design, or in any
area of material science where the properties of the molecules of interest can be reliably predicted.
The designed compounds are generated by applying a set of well-documented organic chemistry reactions to
a database of available starting materials. The outcome of a Synopsis run is a list of good-scoring
compounds together with their proposed synthesis route from these reactions and available starting materials.
An Illustrative Example: Capsaicin look-alikes
To illustrate the working of Synopsis, we have used the program to generate molecules that
resemble capsaicin, known as the molecule that gives chili peppers their spiciness by binding to the
Vanilloid TRPV1 receptor:
Antagonists of this receptor are being investigated as novel agents for pain management.
Syopsis was run with as scoring function the SEAL score (Steric and Electronic Alignment,
S.K. Kearsley and G.M. Smith, Tetrahedron Comput. Method., 3, 1990, 615-633) between
the generated molecules and capsaicin. Below, the structure and electrostatically colored
surface of the best scoring molecule generated in one particular Synopsis run is shown,
together with its synthesis in two reaction steps from commercially available compounds:
FD Computing: Synopsis
Synthesis of Syn009693 in 2 steps:
step 1: reaction = SUZUKI reagent(s):
MFCD00002533
MFCD01009694
product: Int015717
step 2: reaction = AMIDEFROMACID reagent(s):
Int015717
MFCD00144934
product: Syn009693
Two other molecules, generated by different
Synopsis runs are shown below. To view their synthesis routes, click on the 2D structures:
FD Computing: Synopsis
Synthesis of Syn003739 in 1 step:
step 1: reaction = UREA reagent(s):
MFCD00002030
MFCD00040765
product: Syn003739
FD Computing: Synopsis
Synthesis of Syn009902 in 3 steps:
step 1: reaction = ESTER reagent(s):
MFCD00626845
MFCD00082203
product: Int018588
step 2: reaction = BUCHWALDHARTWIG reagent(s):
Int018588
MFCD00239388
product: Int018589
step 3: reaction = REDALDOKETO reagent(s):
Int018589
product: Syn009902
For comparison we show the structure and molecular surface of 3 known synthetic capsaicin
antagonists that are either clinical candidates or lead compounds:
It may be appreciated that even with a very simple scoring function the overall shape and
electrostatics of both capsaicin and its synthetic analogs are reflected in the molecules
generated by Synopsis.
In an actual de novo design effort the scoring function will be a more elaborate and carefully validated
computational protocol derived from structural and biological data on the target system.
Additionaly, ADME-Tox or other desirable properties can be included in the scoring function.
For computational efficiency Synopsis can be run in parallel on a multiple cpu machine.