A logical framework for detecting anomalies in drug resistance algorithms

Virology research is nowadays a discipline involving a broad number of researchers gathered in different institutes and cooperating on defined issues. An example of such an endeavor is the research tackling anti-HIV treatment problems [7] conducted within the Virolab project. The main objective of the ViroLab project is to develop a Virtual Laboratory for Infectious Diseases that facilitates medical knowledge discovery and decision support for HIV drug resistance. Large, high quality in-vitro and clinical patient databases which can be used to relate genotype to drug-susceptibility phenotype have become available. The core of the ViroLab Virtual Laboratory is a rule-based ranking system. More specifically, using a Grid-based service oriented architecture, Virolab vertically integrates the biomedical information from viruses (proteins and mutations), patients and literature (drug resistance experiments), resulting in a rule-based decision support system for drug ranking. This paper is a contribution to virologists, epidemiologists and clinicians in medical knowledge discovery and decision support. The final aim is reasoning on the properties of algorithm modeling the interaction among drugs and HIV virus and detecting its anomalies such as rules that can never be satisfied and subset of rules that are in contradiction.