their activities against HIV. It is also found that the proposed method is capable of
generating structures of known active compound that has scaffold different from
that of the starting one. Furthermore, the structure generation starts from a vertex
which plays a role in predicting biological activity. These observations seem to
address the relationship of the present method [15] with two important aspects of
modern-day drug discovery research—scaffold hopping and inverse QSAR
(iQSAR) problem. Therefore, it appears that this newly developed method [15] may
find useful applications in designing novel therapeutic candidates and may be
helpful for working with drug resistance problems where compounds of very different molecular architecture may be sought for.
Our work presents an interesting alternative to “3D” drug discovery, where
actual molecular coordinates in Cartesian space is used. Combinatorial design and
generation in three-dimensional space would be far more expensive compared to
our approach. Interestingly, one can always follow up on “3D” drug discovery
based on molecule predictions from our method. This would allow a far tractable
approach to drug discovery compared to a seemingly infinite exploration of
molecules in actual “3D” Cartesian space.
Regarding future work, it may be worth exploring whether application of any
quantitative measure for activity prediction can help screen potential bioactive
compounds more effectively. Also, incorporation of new rooted tree-based compound generation and searching algorithms in the existing computer program would
be another important aspect to work on. Finally, it would be of special interest to
see how incorporation of ADME/Tox and drug-able property filters in the computer
program can help discover drug molecules having desired pharmacological and
undesired toxicological activities using the present method.
References
1. Ruddigkeit L, Van deursen R, Blum LC, Reymond JL (2012) Enumeration of 166 billion
organic small molecules in the chemical universe database GDB-17. J Chem Inf Model
52:2864–2875
2. Hansch C, Sammes PG, Taylor JB, Ramsden C (1990) Comprehensive medicinal chemistry:
quantitative drug design, vol 4. Pergamon Press
3. Kier LB, Hall LH (1986) Molecular connectivity in structure-activity analysis. Research
Studies Press
4. Stuper AJ, Brügger WE, Jurs PC (1979) Computer assisted studies of chemical structure and
biological function. Wiley
5. Kitchen DB, Decornez H, Furr JR, Bajorath J (2004) Docking and scoring in virtual screening
for drug discovery: methods and applications. Nat Rev Drug Discov 3:935–949
6. Cramer RD (2003) Topomer CoMFA: a design methodology for rapid lead optimization.
J Med Chem 46:374–389
7. Sun H, Tawa G, Wallqvist A (2012) Classification of scaffold-hopping approaches. Drug
Discovery Today 17:310–324
8. Tanwar J, Das S, Fatima Z, Hameed S (2014) Multidrug resistance: an emerging crisis.
Interdiscip Perspect Infect Dis 2014
Combinatorial Drug Discovery from Activity-Related Substructure …
107
generating structures of known active compound that has scaffold different from
that of the starting one. Furthermore, the structure generation starts from a vertex
which plays a role in predicting biological activity. These observations seem to
address the relationship of the present method [15] with two important aspects of
modern-day drug discovery research—scaffold hopping and inverse QSAR
(iQSAR) problem. Therefore, it appears that this newly developed method [15] may
find useful applications in designing novel therapeutic candidates and may be
helpful for working with drug resistance problems where compounds of very different molecular architecture may be sought for.
Our work presents an interesting alternative to “3D” drug discovery, where
actual molecular coordinates in Cartesian space is used. Combinatorial design and
generation in three-dimensional space would be far more expensive compared to
our approach. Interestingly, one can always follow up on “3D” drug discovery
based on molecule predictions from our method. This would allow a far tractable
approach to drug discovery compared to a seemingly infinite exploration of
molecules in actual “3D” Cartesian space.
Regarding future work, it may be worth exploring whether application of any
quantitative measure for activity prediction can help screen potential bioactive
compounds more effectively. Also, incorporation of new rooted tree-based compound generation and searching algorithms in the existing computer program would
be another important aspect to work on. Finally, it would be of special interest to
see how incorporation of ADME/Tox and drug-able property filters in the computer
program can help discover drug molecules having desired pharmacological and
undesired toxicological activities using the present method.
References
1. Ruddigkeit L, Van deursen R, Blum LC, Reymond JL (2012) Enumeration of 166 billion
organic small molecules in the chemical universe database GDB-17. J Chem Inf Model
52:2864–2875
2. Hansch C, Sammes PG, Taylor JB, Ramsden C (1990) Comprehensive medicinal chemistry:
quantitative drug design, vol 4. Pergamon Press
3. Kier LB, Hall LH (1986) Molecular connectivity in structure-activity analysis. Research
Studies Press
4. Stuper AJ, Brügger WE, Jurs PC (1979) Computer assisted studies of chemical structure and
biological function. Wiley
5. Kitchen DB, Decornez H, Furr JR, Bajorath J (2004) Docking and scoring in virtual screening
for drug discovery: methods and applications. Nat Rev Drug Discov 3:935–949
6. Cramer RD (2003) Topomer CoMFA: a design methodology for rapid lead optimization.
J Med Chem 46:374–389
7. Sun H, Tawa G, Wallqvist A (2012) Classification of scaffold-hopping approaches. Drug
Discovery Today 17:310–324
8. Tanwar J, Das S, Fatima Z, Hameed S (2014) Multidrug resistance: an emerging crisis.
Interdiscip Perspect Infect Dis 2014
Combinatorial Drug Discovery from Activity-Related Substructure …
107