Table 1. Download the benchmark datasets used by MapPred.
No. Data Remark Size MD5
1 Training set List of 6688 sequences in the training set for DeepMeta 40K dfe679a94fd39277bb78fed102f08cc9
2 Training set List of 6289 sequences in the training set for DeepMSA 40K fd3a59c8bd1820468d93b034811008d3
3 Validation set List of 590 sequences in the validation set for all models 4K e3974c0d36855a95ac639a9122ed582d
4 Testing set 41 sequences from Raptrox-Contact (Fasta, MSA, SS, SOLV and PDB) 43M 3df701a5e633a603443f729a712b49be
5 Testing set 38 free modeling domains of 30 targets from CASP12, (Fasta, MSA, SS, SOLV and PDB) 18M 276600bc40783c32482294bd1c9e7744
6 Testing set 228 sequences from Test-Hard of SPOT-Conatct (Fasta, MSA, SS, SOLV and PDB) 97M 3a3a1cd9e81063a73818853a9cb17cab

NOTICE: All tarball files can be uncompressed using the command line: tar jxvf xxx.tar.bz2