This work represents one of the largest screening efforts to date specifically intended to identify and characterize
chemical-assay interference via luciferase inhibition and autofluorescence, and to interrogate the influence of cell
types and culture conditions.
The resulting predictive models can be used to predict interference potential of new chemicals, and to provide
insight into structural features that may influence activity and inform molecular design and assay selection.
Research publication:
High-Throughput Screening to Predict Chemical-Assay Interference, Scientific Reports
Webserver publication:
InterPred: a webtool to predict chemical autofluorescence and luminescence interference, NAR webserver
* This website is free and open to all users and there is no login requirement.
Three assay platforms were applied to analyze luciferase and fluorescence
interference patterns using the Tox21 chemical screening library.
The raw data are freely available on the NCATS Tox21 browser:
All code to calculate the descriptors, build and run QSAR models, and perform analyses detailed in Borrel et al. can be found on github: Source
The QSAR modeling workflow was conducted according to the best practices (Tropsha and Golbraikh, 2007; Cherkasov et al., 2014; Golbraikh et al., 2014). Classification models to predict active versus inactive chemicals for each of the interference assay endpoints were built using Random Forests machine learning. Each model was tuned via a grid optimization and parameters were chosen to maximize performance on a ten-fold cross validation using Matthew’s correlation coefficient (MCC). Model performance was reported as a mean with associated standard deviation on the ten repetitions for the training set, the cross-validation, and the external test set. For details please refer to Borrel et al.
In total we built 17 models to predict interference for HTS assays with respect to luciferase inhibition and autofluorescence under various combinations of color wavelengths, cell cultures and conditions.
Users may select one or several of these models to run on their chemical lists. Please click the button “Submit chemicals for prediction” to begin.