Input features and benchmark data sets for protein complex prediction and E. coli proteome application by AF2Complex

Benchmark data sets of AF2Complex, input features for application to E. coli proteome, and predicted structural models of E. coli Ccm I as described in

Predicting direct physical interactions in multimeric proteins with deep learning

Mu Gao, Davi Nakajima An, Jerry M. Parks, Jeffrey Skolnick

af2complex_bench.tar.gz:  Benchmark data sets CP17, Dimer1193 and Oligomer562, including input features for AF2Complex/AF-Multimer, both paired and unpaired MSAs, as well as sequences, experimental structures, and results presented in the AF2Complex work (~90GB de-compressed size)

ecoli_Ccm_I.tar.gz: Computational models of the E. coli Ccm I system

ecoli_sets.tar.gz: Lists of benchmark sets of positive and negative PPIs from E. coli

ecoli_af_fea.tar.gz: Pre-generated input features of E. coli proteome for protein complex prediction and modeling by AF2Complex (4,429 proteins, ~800 GB de-compressed size). This data set can be used with AF2Complex to probe the interactions of any combinations among the 4,429 proteins of E. coli.