Skip to main content

Table 3 Our method predicted nutrients with an accuracy of 72.5% comparing to 91 experimental data points

From: Computing minimal nutrient sets from metabolic networks via linear constraint solving

Source

Input

Experimental

TP

TN

% Total accuracy

FP

FN

 

nutrients

evidence available

     

Carbon

111

91

30

36

72.5%

17

8

  1. For each element, the table lists how many compounds were predicted to provide that element (Input nutrients), how many of those had experimental evidence against which they could be evaluated (Experimental evidence available, EV), and the results of that evaluation. A compound is a True Positive (TP) if it was predicted to provide that element and did so. A compound is a True Negative (TN) if it was predicted to not provide that element and actually did not. False Positives (FP) and False Negatives (FN) are also reported. The accuracy is obtained by dividing TP + TN by EV and multiplying by 100.