To illustrate the estimation process for the unknown samples, we provide two examples form external validation to demonstrate the applications and effectiveness of the wound age predicted method using the model. In the experiment, we take the skeletal muscle from the rats induced the mechanical injury at P and Q. Then, the expressions of 7 target genes in the two samples were detected respectively(2 - ∆∆ ct ) , as shown in table 4. Then, the sample characteristic expression information of P and Q is represented by form of the model, that is the set of vector combination P={-1,-0,-1,1,1,0,1}, Q={-1,-1,-1,1,1,0,1}. To find a proper forecast.Then, we can calculate the cosine similarity value of (P, N) and (Q, N) (N=4,8,12,16,20,24,28,32,36,40,44,48) by program. The predicted wound age P and Q is derived by the CS value, which we are shown in table 5. P was predicted to be 24h post-injury; Q was predicted to be 36h post-injury, and the actual damage time was 40h(CS value>80%). Table 3 The relative expressions of P and Q as well as the vector values after transformation in the control group. Target gene Pum2 Tab2 Gjc1 Chran1 Abhd2 Mad2l2 Arid5a The expression quantity P(2 - ∆∆ ct ) 0.35 0.72 0.38 4.71 6.36 1.47 1.95 Q(2 - ∆∆ ct ) 0.63 0.65 0.48 2.88 9.79 1.23 1.60 0H(mean ± Std.) 1 ± 0.10 1 ± 0.07 1 ± 0.32 1 ± 0.29 1 ± 0.46 1 ± 0.30 1 ± 0.33 Vector P -1 0 -1 1 1 1 1 Q -1 -1 -1 1 1 0 1 Table 4 The CS values of the unknown sample vector and the model vector matrix. Real/ Prediction(%) 4H 8H 12H 16H 20H 24H 28H 32H 36H 40H 44H 48H P 81.65 81.65 41.14 91.29 83.33 100 57.74 40.82 33.33 30.86 54.77 47.14 Q 61.24 81.65 70.71 73.03 83.33 83.33 57.74 40.82 83.33 92.58 73.03 70.71