According to the paper [1], published in <Bioinformatics>, the prediction accuracy of our PSRSM [2] is rank one. The compared results are shown in the Table 2 of the paper [1].

Reference:

[1] Spencer Krieger, John Kececioglu. Boosting the accuracy of protein secondary structure prediction through nearest neighbor search and method hybridization. Bioinformatics, 36, 2020, i317–i325.

[2] Ma Y, Liu Y, Cheng Y. Protein secondary structure prediction based on data partition and semi-random subspace method. Scientific Reports. 8,9856(2018).

Usages Instructions:

one sequence

1.Submit amino acid sequences with one of the following two ways:

 a) Input sequence with uppercase letters. For example:TIACPHKGCTKMFRDNSAMRKHLHTHG

 b) Upload a fasta file. For example: a.fasta.txt

    >SEQ1

    MTKQEKTALNMARFIRSQTLTLLEKLNELADAADEQADICESLHDHADELYRSCLARFGDDGENL  

2.Waiting for 1-30 minutes according to the sequence length.

Multiple sequences

1.Submit amino acid sequences with one of the following two ways:

 a) Input sequences consist of uppercase letters. For example:

    >SEQ1

   TIACPHKGCTKMFRDNSAMRKHLHTHG

    >SEQ2

   QLCLLCQTSRDCNYIIWTVCRDGCCNIS 

b) Upload a fasta file. For example: a.fasta.txt

    >SEQ1

   TIACPHKGCTKMFRDNSAMRKHLHTHG

    >SEQ2

2.Waiting for 1-30 minutes each sequence according to the sequence length.

8 class of secondary structures->3 class:

GHI →H     BE →E     others→C