Skip to main content
Fig. 2 | BMC Bioinformatics

Fig. 2

From: iMEGES: integrated mental-disorder GEnome score by deep neural network for prioritizing the susceptibility genes for mental disorders in personal genomes

Fig. 2

Correlation between the predictors used in iMEGES on training dataset 1. The feature vector contains different non-coding variants scores, enhancers, promoters and known brain eQTLs for variant prioritization of iMEGES. a Correlation plots illustrate Pearson correlation between each of all scores from the predictors and outcome on the training dataset. The level of correlation was indicated by color and size of the shaded regions in the pie charts at upper right, and blue and larger proportions of the shaded regions suggest higher positive correlation. b Violin plots of scores from various predictors on the training datasets in associated and non-associated groups, and in each violin plot, the median and the first through third interquartile range of predicted scores in each group were provided

Back to article page