ADMET Property Prediction Powered by POEM Available Now
In July 2018, we released a validation note on the novel machine learning technology called POEM. By boosting the latent predictive signals in various chemical fingerprints, POEM outperforms other classifier methods when using the same dataset. Today, we are thrilled to advise that over the past few months our development team has incorporated ADMET Property Prediction powered by the POEM engine into our online platform, Ligand Express®. Several new dashboard tiles and an expanded sidebar menu allow convenient access to Property Prediction featuring our accurate ADMET models. Thanks to high performance computing, an intuitive interface and a parameter-free algorithm, users can easily begin predicting ADMET properties for their chemical libraries. The platform also supports convenient comparison of ADMET properties, with more ways to explore the data already in development.
Read on for some of the key features of Property Prediction method, including the ADMET property collection.
Researchers explore hundreds of molecules during drug development in hopes of ultimately entering the clinic. Cyclica’s online platform makes it easy to submit multiple molecules for the rapid assessment of key molecular properties. Submit molecules by entering their smiles string or upload query compounds as an .smi, .sdf, or .mol2 file. Multiple compounds can also be submitted at once as either a .smi or .csv file with the option of providing molecule names within these files.
Organize Compounds in Libraries
Our flexible library system provides a simple, intuitive way to organize molecules submitted from different projects, and readily reorganize molecules based on property predictions. For example, users can organize the molecules most likely to have desirable properties to a preferred library, with candidates discovered later being easily added to this priority list of compounds
Interact with Results to Identify Leading Candidates
As multiple molecules are added to the platform and run through our predictive models, scientists can readily assess properties using our interactive tools. Easily rearrange heatmaps of data to visualize properties from the most promising candidates. Select molecule details to dive into their details, and view the most similar molecule from each property’s dataset.
Property Prediction on Ligand Express will continue to grow over the months to come. Planned additions include support for customers to upload their custom models and the ongoing expansion of the number of properties offered within this platform. Other ongoing efforts include interactivity improvements to Ligand Express analysis views. Please check back in later this month for some exciting releases as we announce more additions to our platform. Follow our Linkedin page here and our Twitter page here for updates.
For all updates from this month’s release and more, please visit https://cyclicarx.com/product.
Are you a fan of the changes made to Ligand Express? All feedback is greatly appreciated! We would love to hear from you, please send any feedback to firstname.lastname@example.org.
We are looking for some enthusiastic testers to help Cyclica refine its technology. If you are interested please contact email@example.com.