How to Use

Query Sequence(s)

ERPred process only protein sequence file in fasta format. User have option to cut and paste in the text box or directly upload the sequence file. The limits of ERPred is upto 25 sequences at a time. If user submit more than 25 sequences, ERPred predict only first 25 sequences. Following figure 1 is an example showing how to submit a query in ERPred.

SVM Threshold Value

User also have the option to set the threshold. If user select high threshold, the number of prediction will be low but rate of false positive prediction will also be low. If the user select low threshold, the number of prediction will be high and the false positive prediction will also be high.

Example Sequence
>sp|P22071|3BHS1_RAT 3 beta-hydroxysteroid dehydrogenase/Delta 5-->4-isomerase type 1 OS=Rattus norvegicus GN=Hsd3b1 PE=2 SV=3
MPGWSCLVTGAGGFVGQRIIRMLVQEKELQEVRALDKVFRPETKEEFSKLQTKAKVTMLE
GDILDAQYLRRACQGISVVIHTAAVIDVSHVLPRQTILDVNLKGTQNILEACVEASVPAF
IYCSTVDVAGPNSYKKIILNGHEEEHHESTWSDAYPYSKRMAEKAVLAANGSILKNGGTL
HTCALRPMYIYGERSPFLSVMILAALKNKGILNVTGKFSIANPVYVGNVAWAHILAARGL
RDPKKSQNVQGQFYYISDDTPHQSYDDLNCTLSKEWGLRLDSSWSLPLPLLYWLAFLLET
VSFLLRPFYNYRPPFNCHLVTLSNSKFTFSYKKAQRDLGYVPLVSWEEAKQKTSEWIGTL
VEQHRETLDTKSQ
>sp|P27365|3BHS1_MACMU 3 beta-hydroxysteroid dehydrogenase/Delta 5-->4-isomerase type 1 OS=Macaca mulatta GN=HSD3B1 PE=2 SV=2
MTGWSCLVTGAGGFLGQRIVRLLVEEKELKEIRVLDKAFRPELREEFSKLQNKTKLTVLE
GDILDEPFLKRACQDVSVVIHTACIIDVFGVTHRESIMNVNVKGTQLLLEACVQASVPVF
IYTSTLEVAGPNSYKEIIQNGHEEEPLENTWPAPYPYSKKLAEKAVLAANGWTLKNGGTL
YTCALRPMYIYGEGGPFLSASINEALNNNGILSSVGKFSTVNPVYVGNVAWAHILALRAL
RDPKKAPSVQGQFYYISDDTPHQSYDNLNYILSKEFGLCLDSRWSLPLALMYWIGFLLEV
VSFLLSPVYSYQPPFNRHTVTLSNSVFTFSYKKAQRDLAYKPLYSWEEAKQKTVEWVGSL
VDRHKETLKSKTQ
>sp|P36639|8ODP_HUMAN 7,8-dihydro-8-oxoguanine triphosphatase OS=Homo sapiens GN=NUDT1 PE=1 SV=3
MYWSNQITRRLGERVQGFMSGISPQQMGEPEGSWSGKNPGTMGASRLYTLVLVLQPQRVL
LGMKKRGFGAGRWNGFGGKVQEGETIEDGARRELQEESGLTVDALHKVGQIVFEFVGEPE
LMDVHVFCTDSIQGTPVESDEMRPCWFQLDQIPFKDMWPDDSYWFPLLLQKKKFHGYFKF
QGQDTILDYTLREVDTV

Figure 1. Figure A represents ERPred submission page and B represents prediction result pase.

Please cite: Kumar R, Kumari B, Kumar M. Prediction of endoplasmic reticulum resident proteins using fragmented amino acid composition and support vector machine. PeerJ. 2017;5:e3561. doi:10.7717/peerj.3561