Methods for Data Mining from Large Multinational Surveillance Studies

AUTOR(ES)

Poupard, James

FONTE

American Society for Microbiology

RESUMO

Traditionally, large surveillance studies have been analyzed by the use of the MICs at which 90% of isolates tested are inhibited (MIC90s), MIC50s, frequency distributions, and percent susceptibility. In the past, these approaches have proved satisfactory for the monitoring of resistance. From these traditional uses, one can readily detect an increase in MICs for organism and drug combinations. Now that large surveillance studies have been conducted for a number of years and databases have grown to include a large number of datum points, new approaches to the extraction of useful information from these studies are needed. The present study proposes approaches, including the use of antibiotypes, principal components analysis, phylogenetics, and population genetic analysis, to the evaluation of data from large multinational surveillance studies. Application of these types of analyses can be used to describe genetic diversity, analyze changes in susceptibility patterns over time, and possibly, shed light on the origins and evolution of antimicrobial resistance. As global surveillance studies become more common and new questions concerning the evolution of resistance are raised, innovative approaches to analysis of the data will increase in importance.

Documentos Relacionados

• Strong-association-rule mining for large-scale gene-expression data analysis: a case study on human SAGE data
• Association Rules and Data Mining in Hospital Infection Control and Public Health Surveillance
• Practical Considerations for Matching STD and HIV Surveillance Data with Data from Other Sources
• Mining SAGE data allows large-scale, sensitive screening of antisense transcript expression
• Data mining in large sets of complex data