Figure 4
From: Predicting synonymous codon usage and optimizing the heterologous gene for expression in E. coli
The prediction performance of the 18 classifiers for the 18 amino acids with different matched cutoff and window size (A) five amino acids, (B) seven amino acids) in E. coli. The x-axis is the matched percent and the y-axis is the prediction accuracy of the 18 classifiers. Each open circle represents the prediction accuracy with one of the 18 classifiers. The horizontal divisions (from top to bottom) in each box are the upper whisker, 3rd quartile, median, 1st quartile and lower whisker, respectively. The cross line in each box is the mean prediction accuracy of all 18 classifiers. All of the results were calculated based on a ten-fold cross validation.
