Fig. 2
From: SNPeBoT: a tool for predicting transcription factor allele specific binding
Preprocessing stages of data. Each stage is labelled with the number of data points therein and color coded for the presence of classes (gain, loss and no-change). The distribution of classes for train, validation and testing data sets are (gain:2959, loss:2959, no-change:2959), (gain:129, loss:206, no-change:599), (gain:1140, loss:1822, no-change:5438)) respectively