Fig. 5 | Scientific Reports

Fig. 5

From: Integrated machine learning analysis of proteomic and transcriptomic data identifies healing associated targets in diabetic wound repair

Fig. 5

Logistic Regression Model Prediction for (A) Stage I, (B) Stage II, and (C) Stage III in serum measurements of diabetic patients with healing DFUs. The horizontal axis denotes \({Z}^{(i)}={{\varvec{X}}}^{(i)}{\varvec{\beta}}\) and the vertical axis is the probability of being in the faster healing group, i.e., Y(i) = 1. Thus, the dashed lines indicate the patients in the latent space Z and their predicted probabilities of being in the faster healing group. The solid circles denote the ground truth. It is shown that in the latent space, Z, the fast and slow healing groups can be clearly distinguished.

Back to article page