A – Generalized schematic representation of the functions and their relationships in GeneNetwork related to three types of data; gene expression, phenotype and genotype. B-E examples of typical graphical outputs generated by the GeneNetwork. B – Profile of a QTL scan using the interval mapping function. The blue line graph – Likelihood Ratio Statistic (LRS) profile, green and red line graphs – allelic effects (in our case green = Morex, red = Steptoe), yellow bars – confidence intervals determined using 1000 bootstrap tests, red and grey horizontal lines – upper and lower significance LRS thresholds determined by 1000 permutation tests; C – Any pairwise correlation can be visualized as a scatter plot allowing the correlation structure to be determined. In this case, mRNA abundance values (reported by the GeneChip probe set Contig8601_s_at) were plotted against grain yield values from one of the trials. 'N of cases' – number of segregating lines. Pearson's and Spearman's correlation coefficients and associated p-values (P) are shown on the top right corner. Linear regression line is shown in green.; D – Selected correlates can also be visualized as a QTL Cluster map, which is a genetically ordered heat-map representation of the QTLs from multiple traits that were calculated using single marker linkage analysis. Significant QTLs are shown in a different colour from loci that have no association, and allelic effects are shown in contrasting colours (red and blue in key). E – Association network of 10 correlated genes. As a 'seed', mRNA abundance of the HLH DNA-binding protein gene (Contig20506_at), was used. Pearson's correlation coefficient threshold in this case was |0.8|. Line colours show correlation strength (more intense – higher correlation) and whether it is positive (orange – red) or negative (green – blue).
Druka et al. BMC Genetics 2008 9:73 doi:10.1186/1471-2156-9-73