Figure 4 | Scientific Reports

Figure 4

From: Author Correction: Discriminant analysis and binary logistic regression enable more accurate prediction of autism spectrum disorder than principal component analysis

Figure 4

Testing the predictive power of five biomarkers using receiver operating characteristic curve. Areas under the curve (AUC) and p values are indicated. Analysis was performed on ASD (n = 40) and healthy (n = 40) volunteers. PC1: first principal component scores computed in principal component analysis. Disc1: first discriminant scores computed in discriminant analysis. PProb predicted probability computed by binary logistic regression, K plasma potassium, Na plasma sodium, LDH plasma lactate dehydrogenase, GST plasma glutamate S-transferase, MRC1 mitochondrial respiratory chain complex I activity, PC1 the first principal component in principal component analysis. Figure was generated using IBM SPSS Statistics for Windows, Version 27.0, IBM Corp., Armonk, New York, https://www.ibm.com.

Back to article page