Figure 3 | Scientific Reports

Figure 3

From: Gene expression models based on a reference laboratory strain are poor predictors of Mycobacterium tuberculosis complex transcriptional diversity

Figure 3

Comparison of models performance over different data sets. (A) Distribution of the Pearson correlation values obtained when comparing each gene expression measure with its predicted expression. (B) There is one data point for each model. Each dot in a plot is a measure of the gene expression fold-change between Rv0001 and the gene represented by this model. The y-axis corresponds to the predicted fold-change in gene expression while the x-axis corresponds to the measured fold-change. The upper row refers to the models derived from the TFOE data set. The lower row contains the models derived from the ChIP-Seq data set. The values in the left column were calculated when training and predictions were performed with the same data set12. The values in the right column were calculated when different data sets were used for training and predicting.

Back to article page