PMC:4996392 / 6142-7163 JSONTXT

{"project":"2_test","denotations":[{"id":"27600225-23521829-69474967","span":{"begin":243,"end":245},"obj":"23521829"}],"text":"2.2. Triclustering of Genes in Expression Data\nIn order to identify the genes with similar expression profiles over a subset of replicates and a subset of doses of chemical compounds, we have applied an improved version of δ-TRIMAX algorithm [17], called EMOA-δ-TRIMAX (Evolutionary Multi-objective Optimization Algorithm for δ-TRIMAX). It uses a novel Mean Squared Residue (MSR) score as a coherence measure of the resultant triclusters and aims at finding overlapping triclusters from 3D gene expression dataset [22]. The aim is to find large and maximal triclusters, having a MSR score below a certain threshold. In gene expression data, the program thus groups genes according to similarity of their expression levels over multiple doses/time points, as well as samples (i.e., biological replicates). Subsequently, we have identified the genes that are expressed at significantly higher or lower levels at the clustered doses relative to the controls for further analysis using Limma as described above (Section 2.1)."}