Fig. 6: The performance of the four quality assessment methods on all 68 full-length targets, 27 single-chain full-length targets, and 41 full-length targets that are chains of the assembly targets, respectively.
From: Improving AlphaFold2-based protein tertiary structure prediction with MULTICOM in CASP15
A Per-target GDT-TS and B per-target correlation on the server_model_dataset of 68 full-length targets, where the average of per-target GDT-TS or correlation is shown in the box. C Per-target GDT-TS and D per-target correlation on the server_model_dataset of 27 single-chain full-length targets. E Per-target GDT-TS and F per-target correlation on the server_model_dataset of 41 full-length targets. G Per-target GDT-TS and H per-target correlation on the human_model_dataset of 68 full-length targets. I Per-target GDT-TS and J per-target correlation on the human_model_dataset of 27 single-chain full-length targets. K Per-target GDT-TS and L per-target correlation on the human_model_dataset of 41 full-length targets. The vertical line in each box denotes the median per-target value. The red dot in each box denotes the mean per-target value.
