Antisense peptides are used as potential ligands for protein-protein interactions studies. The methods detailed on this website have been developed to allow in silico identification of protein-protein interactions to be carried out using easily available software. The protocol has specifically been written by Dr Nat Milton of Leeds Beckett University and Neurodelta Ltd. The protocol is aimed at undergraduate and postgraduate students plus academic staff for use in non-commercial activities. The main method is based around antisense peptide screening [1] and has been adapted to protein sequence database screening [2].


The attached Protocol uses a Python script based on the methods outlined in [2],[3],[4] to generate antisense peptide sequences from the mRNA for target proteins. The protocol also includes methods using the antisense peptides in BLAST searches to identify potential protein interactions with the target protein. The second part of the protocol details simple methods for 3D modelling and analysis to predict the structure of complexes between the proteins identified from the antisense peptide screening. The method uses the ZDOCK server to predict protein-protein interactions from published protein structure files for the two interacting proteins [5],[6], with the interacting sequences identified in the BLAST search as potential contact residues for the prediction.

An example of the use of these techniques was a study to identify Alzheimer's amyloid-ß (Aß) binding peptides. These were developed based on generating antisense peptides using the messenger RNA (mRNA) sequence for the Aß 1-43 peptide as a template
[3],[4] that would specifically bind the Aß 1-43 peptide. The Aß antisense peptide sequences and data from their generation plus characterisation were the subject of published patent applications [4]. Subsequent studies demonstrated that similar Aß antisense peptides were able to prevent the toxicity of Aß [7], confirming these original observations.