One of the biggest challenges in drug discovery projects is to find new active molecular scaffolds in order to minimize the risk of the projects due to IP restrictions or undesired lead properties. Many methods are tied to the use of pharmacophoric points that limit the exploration to restricted chemical spaces.
The application of 3D field-based molecular alignment and similarity based on newly derived interaction fields provides access to a structure-agnostic new molecular landscape while retaining the target pharmacological profile.
Finding unique, novel and structurally diverse hits with the desired physicochemical properties will help you to screen chemical libraries, claim for new patentability or boost the creativity of your medicinal chemistry teams.
The accessible (and synthesizable) chemical space – on the order of billions of compounds – grows exponentially every day, subsequently increasing the chances of finding new potential hits and lead compounds. However, screening this huge chemical space can be computationally intensive, creating a tradeoff between the search quality and the screening time. Moreover, many generative methods designed to tackle this problem propose structures that are difficult to synthesize.
Pharmacelera’s technology enables screening huge chemical libraries thanks to the use of extremely efficient algorithms in combination with our precise 3D molecular field descriptors. These algorithms not only find original structures but also take into consideration the synthesizability they have.
Combining these algorithms and descriptors prompts to an efficient, accurate, and fast exploration of a huge chemical space while also identifying original and synthesizable compounds.
Precise physicochemical property descriptions such as electrostatic density or the optimization of geometrical parameters of a compound are processes that cannot be simply explained using only classical molecular mechanics methods.
Getting access to these approaches will impact your drug discovery projects, providing a precise atomistic description of molecular processes, such as property calculations and how they impact in the druggability and drug-like properties of your lead compounds.
Pharmacelera develops and applies quantum mechanics methods to study the fundamental properties of small molecules and their impact on the development of a drug. This expertise is reflected in the software solutions and services delivered to our clients.
There are three key elements that influence the quality of machine learning models for drug discovery: data, machine learning algorithms, and molecular descriptors. However, many times much focus is given to the data and the algorithms, while simple fingerprints or molecular properties are used as descriptors.
Pharmacelera’s tools provide unique 3D descriptors derived from the molecular structure of a compound, that can be used in machine learning methods for property prediction. These descriptors include multiple fields of interaction that are relevant to its pharmacological profile and provide a different level of abstraction less tied to the chemical structures of the training set, therefore leading to a better generalization
3D descriptors provide a more realistic representation of molecule properties that can help generating better predictions. Moreover, the usage of 3D fields of interaction as descriptors can lead to a better generalization due to the fact that they are less tied to the chemical structures used to train the model.
Oct
Creation and interpretation of machine learning models for aqueous solubility prediction
M.Su (Pharmacelera) and E. Herrero (Pharmacelera)
May
Screening and Biological Evaluation of Soluble Epoxide Hydrolase Inhibitors: Assessing the Role of Hydrophobicity in the Pharmacophore-Guided Search of Novel Hits
J. Vazquez (Pharmacelera), T. Ginex (Pharmacelera), A. Herrero (Pharmacelera), C. Morisseau (UC Davis), B.D. Hammock (UC Davis) and F. J. Luque (Univ. of Barcelona)
October
Merging Ligand-Based and Structure-Based Methods in Drug Discovery: An Overview of Combined Virtual Screening Approaches
J. Vazquez (Pharmacelera), M. López (AB Science), E. Gibert (Pharmacelera), E. Herrero (Pharmacelera), F. J. Luque (Univ. of Barcelona)
May
Assessing the Performance of Mixed Strategies to Combine Lipophilic Molecular Similarity and Docking in Virtual Screening
J. Vazquez (Pharmacelera), A. Deplano (Pharmacelera), A. Herrero (Pharmacelera), E. Gibert (Pharmacelera), E. Herrero (Pharmacelera), F. J. Luque (Univ. of Barcelona)
February
Lipophilicity in Drug Design: an Overview of Lipophilicity Descriptors in 3D-QSAR Studies
T. Ginex (Univ. of Barcelona), J. Vazquez (Pharmacelera), E. Gibert (Pharmacelera), E. Herrero (Pharmacelera), F. J. Luque (Univ. of Barcelona)
July
Development and Validation of Molecular Overlays Derived From 3D Hydrophobic Similarity with PharmScreen
J. Vázquez (Pharmacelera), A. Deplano (Pharmacelera), A. Herrero (Pharmacelera), T. Ginex (Univ. of Barcelona), E. Gibert (Pharmacelera), O. Rabal (CIMA), J. Oyarzabal (CIMA), E. Herrero (Pharmacelera), F. J. Luque (Univ. of Barcelona)
July
Application of the Quantum Mechanical IEF/PCM-MST Hydrophobic Descriptors to Selectivity in Ligand Binding
T. Ginex (Univ. of Barcelona), J. Muñoz-Muriedas (GSK), E. Herrero (Pharmacelera), E. Gibert (Pharmacelera), P. Cozzini (Univ. of Parma), F. J. Luque (Univ. of Barcelona)
January
Development and validation of hydrophobic molecular fields from the quantum mechanical IEF/PCM-MST solvation models in 3D-QSAR
T. Ginex (Univ. of Barcelona), J. Muñoz-Muriedas (GSK), E. Herrero (Pharmacelera), E. Gibert (Pharmacelera), P. Cozzini (Univ. of Parma), F. J. Luque (Univ. of Barcelona)
