Chemical space and compound selection

Visualize the distribution of properties and scores across the chemical space of your project to identify chemistries most likely to succeed. Select sets of compounds with the desired balance between quality and overall diversity to spread risk as appropriate across multiple chemistries.

• Create a chemical space projection based on molecular structure (‘chemical structure’ space plot), the properties in the data set (‘property’ space plot) or a combination of the two (‘mixed’ space plot)

• Plot multiple data sets in one chemical space projection - a useful tool when comparing different chemical series within a project

• View the structure of each compound in the plot by pointing at the corresponding point

• Select compounds from the chemical space plots for more detailed inspection or progression to further analysis

• Visualize trends by coloring the dots in the chemical space plot on a user-defined scale according to any property or probabilistic score

While a scoring scheme, such as probabilistic scoring, allows the ‘quality’ of the compounds in the library to be assessed against a project’s target profile across a broad range of properties, it may not be appropriate to choose only the ‘best’ compounds from the library for progression. In particular, early in a project, consideration should be given to exploring a diversity of chemistry. The purpose of this is to provide additional structure-activity data, validation of the accuracy of property predictions and, if possible, multiple chemical series to provide alternatives should an unexpected difficulty arise with the lead chemical series.

In combination with the chemical space plots and probabilistic scoring, the compound selection algorithm provides the ability to explore the impact of different compound selection strategies on the likelihood of success. There may be a good argument for manual selection of specific compounds, to test a specific hypothesis or even out of curiosity, and this should not be ruled out. However, the objective balancing of diversity and quality encourages the majority of research effort to be expended in those areas of chemistry most likely to yield progress toward a project’s objectives.

Download our Fequently Asked Questions documents on Chemical Space or Compound Selection for further information.

StarDrop™, Glowing Molecule™ and Auto-Modeler™ are trademarks of Galapagos NV and/or its affiliates


search Target discovery Hit finding biology Hit finding libraries Hit-to-lead/Lead optimization Predictive drug discovery Chemogenomics StarDrop - Identify balanced compounds - Focusing resources - Data integration - Demos and evaluations - Licensing - Overview · ADME QSAR models · P450 models · Glowing Molecule · Chemical space · Probabilistic scoring · Data analysis · Auto-Modeler Compound management Partnerships About us News and events Careers Contact BioFocus DPI Home Galapagos Home Mailing list	Visualize the distribution of properties and scores across the chemical space of your project to identify chemistries most likely to succeed. Select sets of compounds with the desired balance between quality and overall diversity to spread risk as appropriate across multiple chemistries. • Create a chemical space projection based on molecular structure (‘chemical structure’ space plot), the properties in the data set (‘property’ space plot) or a combination of the two (‘mixed’ space plot) • Plot multiple data sets in one chemical space projection - a useful tool when comparing different chemical series within a project • View the structure of each compound in the plot by pointing at the corresponding point • Select compounds from the chemical space plots for more detailed inspection or progression to further analysis • Visualize trends by coloring the dots in the chemical space plot on a user-defined scale according to any property or probabilistic score While a scoring scheme, such as probabilistic scoring, allows the ‘quality’ of the compounds in the library to be assessed against a project’s target profile across a broad range of properties, it may not be appropriate to choose only the ‘best’ compounds from the library for progression. In particular, early in a project, consideration should be given to exploring a diversity of chemistry. The purpose of this is to provide additional structure-activity data, validation of the accuracy of property predictions and, if possible, multiple chemical series to provide alternatives should an unexpected difficulty arise with the lead chemical series. In combination with the chemical space plots and probabilistic scoring, the compound selection algorithm provides the ability to explore the impact of different compound selection strategies on the likelihood of success. There may be a good argument for manual selection of specific compounds, to test a specific hypothesis or even out of curiosity, and this should not be ruled out. However, the objective balancing of diversity and quality encourages the majority of research effort to be expended in those areas of chemistry most likely to yield progress toward a project’s objectives. Download our Fequently Asked Questions documents on Chemical Space or Compound Selection for further information. StarDrop™, Glowing Molecule™ and Auto-Modeler™ are trademarks of Galapagos NV and/or its affiliates	Related links Demos and Evaluations Case Studies Licensing ADME/PK Laboratory Contact us BioFocus DPI Home Galapagos Home