Fig. 5: Selecting accurate conformations.
From: Structure prediction of alternative protein conformations
a TM-score vs plDDT. The Pearson correlation between the TM-score and plDDT is low (0.52), suggesting that the quality metric can’t be used to distinguish between alternative conformations. b TM-score to conformations in the training set vs. TM-score to unseen conformations using the best strategy (MSA clustering) coloured by plDDT. Only structures that could be predicted with a TM-score >0.8 for both train and test conformations among the samples taken are displayed (52 structures, n = 5408 sampled predictions). c Examples of predicted conformations coloured by plDDT/grey and in structural superposition. Some predictions are less accurate than others, resulting in chain breaks and low plDDT regions (orange).
