CellPred provides a web service for predicting cell percentage based on high throughput gene expression data (e.g. microarray data).
- Analysis based on files CellPred directly reads uploaded data files.
- Online computation Computation will be done on the server. Users only need a web browser to upload files, to submit requests for analysis and to view results.
- User-specified factors The user is allowed to add variables (e.g., batch effect) into the multivariate model for prediction.
- Modified quantile normalization The modified quantile normalization uses training data set as the reference to normalize the test set.
- Cell percentage analysis Multiple linear regression models are used for prediction.
Steps for analysis:
- (optional) logon/register
- upload files
- set parameters for model-training for percentage-prediction analysis and submit requests
- browse/download results