Fig. 6: Cross-examination of class I and II models against class I and II data.
From: Catch bond models may explain how force amplifies TCR signaling and antigen discrimination
a, b \({\delta }_{0}^{*}\) vs \(I\) plots (data presented with the best-fitting parameters with SE) obtained using class І (black) or class II (red) model to fit force-lifetime data of TCR interacting with pMHC-І (a) or pMHC-II (b) molecules. In each panel, two sets of parameter values were returned from fitting depending on whether class I (black) or class II (red) model was used because they are based on different structures and conformational changes of the TCR–pMHC complexes. c The slopes (gray, the level of correlation between \({\delta }_{0}^{*}\) and \(I\)) and goodness-of-fit (R2) (blue, the degree of appropriateness of the model for the data) of the linear fit in (a, b) are shown in the matched (1st and 4th groups) and mismatched (2nd and 3rd groups) cases. All error bars represent SE of linear fitting. d, e \({\delta }_{0}^{*}\) vs \({n}^{*}\) plots (data presented with the best-fitting parameters with SE) obtained using class І (black) or class II (red) model to fit force-lifetime data of TCR interacting with pMHC-І (d) or pMHC-II (e) molecules. f The slopes and goodness-of-fit of the linear fit in (d, e) are shown in the matched (1st and 4th groups) and mismatched (2nd and 3rd groups) cases. The slopes indicate the average unfolding extension per amino acid (nm/a.a.) from each model, which are compared to the maximum average contour length per amino acid of ~0.4 nm/a.a. (biophysical limit, black dashed line with considerable deviation)19,39,40. All error bars represent SE of linear fitting. Source data are provided as a Source Data file.
