AGP calculates the individual probability of association for each gene with Age Related Disorders(ARDs) in humans. It is a machine learning based tool. The prediction was done on the basis of random forest classifier. The distinguishing parameters of ARDs and NON-ARDs are physical properties of amino acid sequences of corresponding genes, sub-cellular localization and network topologies.
AGP has precision and sensitivity of 75% and 76% respectively. Area under the receiver operating characteristic curve is 85% suggesting high accuracy of the predictions.
You can search for gene with query tool, using any of the following gene identifiers : HGNC Gene Symbol, Entrez ID, Uniprot Swissprot Accession ID.