Data and most relevant results for the FoldDock project

Data and main results for the study "Improved prediction of protein-protein interactions using AlphaFold2 and extended multiple-sequence alignments."
Contained datasets consist of:1 - 219 heterodimers from dockground benchmark 4 dataset 2 - 1503 heterodimeric structures from a recent study (Green, A. G. et al. Nat. Commun. 12, 1–12 (2021))3 - 7 heterodimeric complexes from CASP144 - 8 novel heterodimeric complexes deposited in the PDB database after 15 June 2021
For each one of the mentioned datasets it can be found in this repository:- concatenated protein sequences in FASTA format;- 3D structure atomic coordinates in PDB format;
- csv files summarizing chain IDs, chain lengths and other dataset parameters;- msas for pair of sequences, required to run AlphaFold2

For dataset 1 and 2, AlphaFold2 output docking results for 5 identical runs are provided as well, together with a csv summarizing dockq scores and different model quality metrics.