Overview
Multi-feature Relatedness (MFR) model is a machine learning model based on SVM and takes a total of 12 similarity-based gene pair features including both expression similarities and prior-knowledge based similarities as the input. And the output value is applied as an assessment criterion, namely “Multi-Feature Relatedness”, for measuring the relatedness between a pair of genes under given certain conditions (see Figure 1).
Figure 1. Construction of MFR model
From this website, you can: