Fig. 6
From: SwarmTCR: a computational approach to predict the specificity of T cell receptors
Data allocation. The allocation of data for both BS (A) and SC (B) model is illustrated above. All TCRs are split proportionally into the sample and reference sets. As described in the literature, the same SC reference set is used in training and testing (due to abundance of data, this was not necessary in the BS analysis). The subset labeled peptide specific TCRs are the group of TCRs specific for the peptide tested by SwarmTCR