The FoldRec-C2C are implemented in Ubuntu 18.04, python3.6.9, while the plotted figure 4 in manuscript depands on windows 10 All the files are the zipped package. Source: The FoldRec-C2C-V1 contains all the source code of FoldRec-C2C predictor. There are three re-ranking model in FoldRec-C2C we take them out and put them in this site. The seq-to-seq-V1, the seq-to-cluster-V1 and the cluster-to-cluster-V1 are those models source code. Features: The benchmark dataset is to evaluate the performance of FoldRec-C2C. We used HHSearch, Top-1-gram, Top-2-grams, 84-features and DeepFR theses features to generate Learning to Rank (LTR) model after training procession. We assemble these features as the training data and these data the input of RankLib-2.13.jar in the training step. The generated LTR model are put in this site The related software of FoldRec-C2C are HHsuite, Pse-in-one2.0 and RankLib-2.13. Result: The ranking result of three ranking result are put in this site. The result of C2C model is the final ranking result. Evaluate: The data to plot specificity-sensitivity curve of all the method used for comparison. These method are HMMER, THREADER,FOLDpro, RF-Fold, DN-FoldS, DN-Fold, RFDN-Fold, DeepFR, DeepFRpro, FoldRec-C2C(S2S), FoldRec-C2C(S2C) and FoldRec-C2C(C2C). We packaged a folder which contains those file. The source code to plot this specificity-sensitivity curve is contained in the FoldRec-C2C-V1 package