Domain-enhanced analysis of microarray data using GO annotations

MOTIVATION: New biological systems technologies give scientists the ability to measure thousands of bio-molecules including genes, proteins, lipids and metabolites. We use domain knowledge, e.g. the Gene Ontology, to guide analysis of such data. By focusing on domain-aggregated results at, say the molecular function level, increased interpretability is available to biological scientists beyond what is possible if results are presented at the gene level. RESULTS: We use a 'top-down' approach to perform domain aggregation by first combining gene expressions before testing for differentially expressed patterns. This is in contrast to the more standard 'bottom-up' approach, where genes are first tested individually then aggregated by domain knowledge. The benefits are greater sensitivity for detecting signals. Our method, domain-enhanced analysis (DEA) is assessed and compared to other methods using simulation studies and analysis of two publicly available leukemia data sets. AVAILABILITY: Our DEA method uses functions available in R (http://www.r-project.org/) and SAS (http://www.sas.com/). The two experimental data sets used in our analysis are available in R as Bioconductor packages, 'ALL' and 'golubEsets' (http://www.bioconductor.org/). SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.     

Statistical decoding of potent pools based on chemical structure

Pooling experiments are used as a cost-effective approach for screening chemical compounds as part of the drug discovery process in pharmaceutical companies. When a biologically potent pool is found, the goal is to decode the pool, i.e., to determine which of the individual compounds are potent. We propose augmenting the data on pooled testing with information on the chemical structure of compounds in order to complete the decoding process. This proposal is based on the well-known relationship between biological potency of a compound and its chemical structure. Application to real data from a drug discovery process at GlaxoSmithKline reveals a 100% increase in hit rate, namely, the number of potent compounds identified divided by the number of tests required.