Predictive Toxicology Knowledge Representation and Processing Tools Based on a Hybrid Intelligent Systems Approach








Abstract: The increasing amount and complexity of data used in toxicity prediction calls for new and flexible approaches based on hybrid intelligent methods to mine the data. This research effort is required even more in the context of the increasing number of different classifiers applied in toxicity prediction. The main aim of this project proposal is to explore knowledge representation issues regarding integration of artificial intelligence (Al) structures to propose and validate efficient and robust models to predict, mine and exploit toxicological data. The proposed structures will be based on Hybrid Intelligent Systems (HIS) and Multi-Classifier Systems (MCS) paradigms to model and predict toxicity of chemical compounds. Knowledge discovery techniques will be applied to successfully developed models to mine the toxicity data further. The project proposes an integrated Al platform to combine in silico toxicity profiling of chemical compounds, and experimental validation. Therefore, experts can add results from their experiments and also their expertise, and general users can retrieve data and knowledge that will provide clues to improve the existing toxicity predictive models. To achieve this goal, experimental results and human expertise will be integrated in ensemble structures of connectionist and fuzzy inference systems, thereby selecting useful molecular descriptors and multiple predictive models to produce a single composite model that generally exhibits greater predictive power and generality.

Principal Investigator: Dr CD Neagu

Co-investigator: Dr MTD Cronin

Research Assistant: Dr G Guo

Starts: 1 Oct. 2004

Ends: 30 Sept. 2006

Value: 100,072

Award Type: First Grant Scheme

EPSRC Research Topic Classifications:

Information and Knowledge Management

EPSRC Industrial Sector Classifications: Software

EPSRC Grant Reference: GR/T02508/01

What's New

1. 1st September, 2006, the latest project meeting was held in CSL, York. Project review and further work.

2. 9th August, 2006, a project meeting was held in University of Bradford. Discuss a collaborated paper for ALTA.

3. 6th April 2006, a project meeting with our colleagues from CSL York to discuss on the database design outlines.

4. 21th March 2005, a project meeting with Mark in Liverpool John Moore Univ. To discuss achievement to date and further work.

5. 20th October, 2004, the first project meeting was held in CSL, York. To discuss the data collection and organization.

6. 21st March, 2005, the second project meeting was held in Liverpool John Moores University. To discuss some issues related to data quality and quantity.