Mining large heterogeneous data sets in drug discovery.

Area of Science:

• Computational Chemistry
• Bioinformatics
• Pharmacology

Background:

• Effective drug discovery necessitates searching and data mining of extensive information from diverse sources, including chemistry, biology, and pharmacology.
• The proliferation of databases relevant to drug discovery presents both opportunities and challenges for researchers.

Purpose of the Study:

• To review publicly available large-scale databases pertinent to drug discovery.
• To describe data mining approaches applicable to these databases.
• To discuss recent advancements in integrative data mining, including Semantic Web techniques, for identifying cross-source associations.

Main Methods:

• Literature review of publicly available drug discovery databases.
• Analysis of data mining techniques for large-scale datasets.
• Exploration of integrative data mining strategies and Semantic Web applications.

Main Results:

• Identification of key large-scale databases for drug discovery.
• Overview of various data mining methodologies.
• Demonstration of integrative approaches for multi-source data analysis.

Conclusions:

• Future drug discovery data mining requires intelligent, semantic aggregation of information.
• Advanced methods like intelligent agents and inference engines are essential for client applications.