genes, compounds, pathways that lead to interesting correlations.
Analysis of gene expression in tobacco indicated that the lack of elevated B-6 content was not due to lack of enzymatic activity, but to down-regulation of the endogenous tobacco genes compounded with limited transgene expression.
For instance, in order to find the possible in-directed linkage for a given gene-compound pair, we can look up the four extended associations in Chem2Bio2RDF: gene-disease-compound, gene-compound-compound, gene-pathway-compound, and gene-gene-compound, and then compute the association scores for the outputs from Chem2Bio2RDF.
In Chem2Bio2RDF, there are eight kinds of relations: compound-compound, compound-gene, compound-disease, compound-side effect, compound-pathway, gene-gene, gene-disease, and gene-pathway.
While previous applications of LDA in the biomedical domain have yielded several benefits, few considered the extension of the LDA model to include bio-terms (that is a restricted vocabulary of genes, compounds, diseases, and so on) as input parameters.
As a second estimate of the error rate of the gene-compound correlates, we determined a false discovery rate, as estimated by a q-value, for each correlate.
However, broad gene expression patterns as identified by principle component analysis contribute only moderately to gene-compound relationships in this study.
Simple, fast reports depend on directly related information e.g. a gene report which brings together Sequence Variant, gene-disease and gene-compound information.
Using the fifty-nine celinesnes for discovery and an additional seven cell lines for which extensive compound sensitivity data were available as a test set, we determined that gene-compound pairs with a correlation coefficient above 0.6 had a false discovery rate of approximately 5%.
For these compounds and proteins we use the network search algorithm to search the entire COPD knowledge network but restrict the allowed connections to human genetic interaction, protein-protein interaction, gene-compound or gene-disease interaction (see Table 2 for the individual sources mapped to each of these relation types).
