Front Cover: The cover image has been created by Clemens Meyer and Adam Cain from DeepMind. It shows the ground truth experimental structure for CASP14 target T1044 in green, together with the prediction of the AlphaFold2 team in blue.Target T1044 (PDB 6VR4) is the RNA polymerase of the crAss-like phage phi14:2, a large single-chain protein with 2180 residues. Its structure was published by Drobysheva et al. (Nature, 2020). Target T1044 was split into 9 domains for the CASP14 experiment (T1031, T1033, T1035, T1037, T1039, T1040, T1041, T1042, and T1043); a prediction for the full protein chain, including domain packing, is shown here.

Reverse Cover: The cover image is based on the review Critical assessment of methods of protein structure prediction (CASP) – Round XIV by Andriy Kryshtafovych, Torsten Schwede, Maya Topf, Krzysztof Fidelis and John Moult, DOI 10.1002/prot.26237. The highlight of the CASP14 experiment was an outstanding performance of AlphaFold2 structure modeling method. The cover image shows the backbone agreement with experiment of the best models in CASP1 (1994), CASP5 (2002) and the latest three CASP experiments. The data are shown as linear trend lines; for the most recent round, individual target points are shown as circles. Performance in CASP14 (top black line) is very impressive, with accuracy approaching experimental accuracy for many targets (GDT_TS of 90 or more). This outstanding performance is dominated by AlphaFold2 models, although other groups also advanced substantially from CASP13 (dashed black line). The agreement metric, GDT_TS, is a multi-scale indicator of the closeness of the Cα atoms in a model to those in the corresponding experimental structure. Target difficulty is based on sequence and structure similarity to other proteins with known experimental structures.