AMPZero: Antimicrobial Activity Predictor for Microbial Species


AMPZero is a predictor designed for predicting the effectiveness of a peptide sequence against various microbial species. The predictor takes the peptide sequence and any N/C-termini modifications together with the genomic sequence of a microbial species to generate targeted predictions. For more details related to this work and for citing us please visit - " Sadaf Gull and Fayyaz Minhas, AMP0: Species-Specific Prediction of Anti-microbial Peptides using Zero and Few Shot Learning,” arXiv preprint arXiv:1911.06106, 2019".
AMP0 is a machine learning model which uses zero shot learing to generate predictions about peptide's effectiveness against microbial species. The data we have used for training can be downloaded from here which includes a peptide ID, C-terminus modification, N-terminus modification, Peptide Sequence and MIC value.

  • Webserver provide users with list of C/N termini modifications (OPTIONAL)
  • If user wants to predict peptide's effectiveness for some species then provide their genomic sequences in FASTA format (OPTIONAL).
    This field is optional if user do not upload any file then the results for 336 microbial species will be generated on which model is trained
  • User can input one or more peptide sequences in the given text area in FASTA format (MANDATORY).
  • After finding your required settings click on "Predict" button

Select C-terminus modification:  Select N-terminus modification: 

Upload fasta file of genome sequences (OPTIONAL): 
100 MiB maximum size   Download a test genome file

Peptide Sequences (FASTA format):                  

Developed and maintained by Biomedical Informatics Research Laboratory , Department of Computer and Information Sciences   (DCIS),
Pakistan Institute of Engineering and Applied Sciences  (PIEAS) , Islamabad, Pakistan.