Age Predictor is an online tool to estimate methylation age of the sample by three machine learning methods (SVM, Random Forest and Elastic Net). Users can input the idat file, processed data or sample ID of GEO datasets to obtain the prediction results.
Age Predictor accepts compressed raw data files (.gz format), which contain two files (one for green signal intensity and another for red signal intensity). The Age Predictor will process the raw data using standard Illumina pipeline and return a predicted age.
Age Predictor also accepts processed data. The processed data file should be a tab (\t) delimited text file. The first column of processed data must be CpG probe identifiers (cg numbers), such as cg00000165. The second column of processed data must be beta values (range from 0 to 1). The file contains at least specific 52 probes, which are used in our model (list in probe_list.txt).Users can select the normalization method in the drop-down box.
Users can also specify a NCBI GEO sample ID for age prediction, if its platform is Illumina 450K. Age Predictor first fetches data from NCBI and then estimates the DNA methylation age.