Skip to main content
Fig. 6 | BMC Bioinformatics

Fig. 6

From: Convert your favorite protein modeling program into a mutation predictor: “MODICT”

Fig. 6

MODICT. SCORES FOR PARTIALLY DELETERIOUS PAH MUTATIONS. Top Left. Mutations with residual activity in PAH with their respective MODICT scores are plotted. Triangles indicate data points that are 2 standard deviations apart from the mean (both residual activity and MODICT score) of rectangle data points. Top Right. Outliers that are two standard deviations apart from the mean are removed and the correlation coefficient is calculated. MODICT scores are negatively correlated with residual activity (r=−0.494). The exact p-value of the correlation coefficient is 0.036 based on 1-tailed t-distribution. Middle Left. The same comparison was applied to POLYPHEN2 scores. Triangle data points indicate the outliers. Middle Right. Likewise, POLYPHEN2 scores were negatively correlated with residual activity (r=−0.417). However, the exact p-value of the correlation coefficient was 0.062 based on 1-tailed t-distribution. Lower Left. The training module of MODICT were used on the same mutations. Lower Right. The training module of MODICT was able to achieve a weight score configuration that yielded a more significant p-value of 0.002. (=p<0.1;=p<0.05)

Back to article page