CDD’s Scientific Relations Consultant, Dr. Adam Kallel, is scheduled to present a poster at the Drug Discovery Chemistry Conference in San Diego between April 23rd & 25th. Dr. Kallel’s poster is titled “Leveraging CDD Vault®’s Kinome Data to Build Predictive Models of Kinase Activity”.
Click here to register.
Title: Leveraging CDD Vault®’s Kinome Data to Build Predictive Models of Kinase Activity
Abstract: CDD Vault™ contains hand curated data for over 500 kinases. A random selection of 9 of these kinases were selected and Bayesian binary QSAR models were constructed for each of these kinases. The average accuracy on active compounds was 62.7%. The average accuracy on inactive compounds was 94.2%. The p-values for these models ranged from e-8 to e-52 indicating high significance. We conclude that it is possible to develop a “kinase fingerprint” that is rapidly calculated for compounds registered into CDD Vault™ either for characterization or cross-reactivity checking purposes; however, due to lack of sufficient compounds, some models can be over determined and should be exclude from predictive modeling.
Author: Dr. Adam Kallel, Scientific Relations Consultant, Collaborative Drug Discovery
This blog is authored by members of the CDD Vault community. CDD Vault is a hosted drug discovery informatics platform that securely manages both private and external biological and chemical data. It provides core functionality including chemical registration, structure activity relationship, chemical inventory, and electronic lab notebook capabilities!
CDD Vault: Drug Discovery Informatics your whole project team will embrace!
Translated with Google Translate