 |
Intelligent Pharma scientists and engineers are actively researching and designing new techniques for computational chemistry and artificial intelligence applications. So as a result, Intelligent Pharma has developed computational solutions ready for direct delivery to our customers. In addition, Intelligent Pharma's engineers can customize the developed solutions to the specific needs of every customer.
Products
 |
Helios: A ligand-based virtual screening solution.
The new HELIOS technology is based on the superimposition of Molecular Electrostatic Potentials (MEP) between any compound and a reference active ligand. While working under a Molecular Mechanics force fields, HELIOS not only considers the MEP generated by formal charges that occur in the center of atoms, but also other types arising from the dynamic behavior of molecular electronic clouds.
Helios leaflet- White paper about Helios (it includes methods, comparisons, etc.): sent upon request at ibelda (at) intelligentpharma (dot) com
 |
Selene: A receptor-based virtual screening solution.
The core of SELENE VS is based on docking techniques. Currently, SELENE is using AutoDock 4.0, although it can easily be customized to interface with any other docking program specified by the costumer. Any customized solution can be developed - on demand - by Intelligent Pharma engineers and integrated in the customer's current information technology system. SELENE automatically distributes the docking calculations through an specified "cloud computer", manage the docking results and introduces them in a database, and provides a wide set of statistical tools to help the scientists to analyze the docking results of thousands of compounds.
- Selene leaflet
 |
Chiron: Virtual combinatorial chemistry.
CHIRON is a web-based system to aid in the design and optimization of combinatorial chemistry libraries. Combinatorial chemistry often implies an exponential increase of the expenses without an exponential increase of the productivity. Hence, Chiron has been developed to minimize the number of combinatorial synthesis while keeping the scientific productivity. CHIRON is based on an artificial intelligence system (genetic algorithms) that proposes a reduced subset of the combinatorial library to be synthesized and to be experimentally screened. The results are reintroduced in Chiron and, then, it proposes a second generation of compounds (a second subset of the combinatorial library). This work-flow iterates until a certain optimization level is reached. At Intelligent Pharma, we have proved that by synthesizing and evaluating only between the 10% and the 15% of the combinatorial library, one can reach the same results than synthesizing and evaluating the full combinatorial library. Therefore, Chiron can save up to the 90% of the costs of a combinatorial chemistry program.
 |
Medea: Medea is a data mining solution applied to computational toxicology.
MEDEA uses an artificial intelligence system based on evolutionary computation. Once Medea is trained, it is able to predict with a high degree of precision the toxicology of new compounds. In addition, Medea is also able to explain the reasons for the predictions, due to the data mining platform on which it is based. Medea uses an evolutionary algorithm that evolves rules that model physico-chemical and structural properties of a training set. Therefore, at the end of the evolutionary process, one can analyze the evolved rules and obtain the common properties of toxic and non-toxic compounds. Therefore, using this strategy, medicinal chemists can avoid common toxic properties in further molecular design rounds. This strategy is known as knowledge extraction.
- I. Belda, I. Traus, S. Gordo, T. Tarragó, S. Madurga, X. Llorà and E. Giralt. Peptide data mining: from virtual design to knowledge extraction. Proceedings of the 8th annual conference on Genetic and evolutionary computation: 297 - 298 (2006).
ENPDA: ENPDA is a de novo automated drug design solution. In ENPDA drug design process, the user specifies the protein receptor surface patch where ligands are going to be designed. Therefore, using an evolutionary computation optimization kernel, ENPDA designs new molecules that bind to the specified protein receptor surface patch. If the initial evolutionary population is randomly initialized, the process is known as de novo drug design, otherwise, if the population is initialized based on a given scaffold, the process is known as virtual combinatorial chemistry.
- I. Belda, X. Llorà and E. Giralt. Evolutionary algorithms and de novo peptide design. Soft Comput. 10(4): 295-304 (2006)
- I. Belda, S. Madurga, X. Llorà, M. Martinell, T. Tarragó, M. G. Piqueras, E. Nicolá and E. Giralt. ENPDA: an evolutionary structure-based de novo peptide design algorithm. J. Comput. Aided. Mol. Des. 19(8): 585-601 (2005)