This server is designed for predicting mammalian secreted proteins (SPs). It will generate an email message containing numeric score for each protein in your submission that quantifies putative propensity to be secreted. Putative annotations (a given protein is predicted either as a SP or non-SP) displayed on the result page are based on the judgement threshold input in this page. All comments, suggestions or requests regarding our work are welcome. For more information please contact biocomputinglab@sina.com.

Please follow the four steps below to make a prediction

STEP ONE Paste your protein sequence with standard FASTA format:

  Click Here to Use Example Protein Sequence

STEP TWO Select a prediction model:

STEP THREE Set a judgement threshold:

STEP FOUR Provide you email address:

HELP

iMSP can accept five protein sequences for a sigle prediction. You should input the protein sequence in standard FASTA format, which begins with a symbol ">". The format of the input protein sequence is as follows:

    Line 1: > PDB ID or UniProtKB ID or any other annotation(s)

    Line 2: protein sequence (1-letter amino acid encoding)

    Line 3: > PDB ID or UniProtKB ID or any other annotation(s)

    Line 4: protein sequence (1-letter amino acid encoding)

    Line 5: ...

The judgement threshold that input in STEP THREE will be used to generate putative annotations. Proteins gaining numeric score higher the threshold will be regard as a putative SP.

We will send a message reporting the detailed prediction result to the email address you input once the current prediction is finished. For more information in respect to its content, please make a try and see the result.

Materials

SPs-all: dataset used for Mammalia SPs Download

SPs-H: dataset used for Homo sapiens SPs Download

SPs-M: dataset used for Mus musculus SPs Download

SPs-B: dataset used for Bos taurus SPs Download

SPs-C: dataset used for Canis lupus familiaris SPs Download

SPs-O: dataset used for Oryctolagus cuniculus SPs Download

Citation

Upon the usage the users are requested to use the following citation:

Jian Zhang, Haiting Chai, Song Guo, Huaping Guo, Yanling Li, 20XX, High-throughput Identification of Mammalian Secreted Proteins Using Species-specific Scheme and Application to Human Proteome, XXX, DOI: XXX, PMID: XXX link

Acknowledgments

We acknowledge with thanks the following databases and software used as a part of this server:

UniProtKB - UniProt Knowledgebase

LIBSVM - A Library for Support Vector Machines