Details

Autor(en) / Beteiligte

• Porto, William F.
• Fensterseifer, Isabel C.M.
• Ribeiro, Suzana M.
• Franco, Octavio L.

Titel

Joker: An algorithm to insert patterns into sequences for designing antimicrobial peptides

Ist Teil von

• Biochimica et biophysica acta. General subjects, 2018-09, Vol.1862 (9), p.2043-2052

Ort / Verlag

Netherlands: Elsevier B.V

Erscheinungsjahr

2018

Quelle

MEDLINE

Beschreibungen/Notizen

• Innovative alternatives to control bacterial infections are need due to bacterial resistance rise. Antimicrobial peptides (AMPs) have been considered as the new generation of antimicrobial agents. Based on the fact that AMPs are sequence-dependent, a linguistic model for designing AMPs was previously developed, considering AMPs as a formal language with a grammar (patterns or motifs) and a vocabulary (amino acids). Albeit promising, that model has been poorly exploited mainly because thousands of sequences need to be generated, and the outcome has high similarity to already known AMPs. Here we present Joker, an innovative algorithm that improves the application of the linguistic model for rational design of antimicrobial peptides. We modelled the AMPs as a card game, where Joker combines the cards in the hand (patterns) with the cards in the table (sequence templates), generating a few variants. Our algorithm is capable of improving existing AMPs or even creating new AMPs from inactive peptides. A standalone version of Joker is available for download at <http://github.com/williamfp7/Joker> and requires a Linux 32-bit machine. [Display omitted] •Joker performs modifications on peptide sequences in a sliding window fashion.•Joker requires hundreds or only tens of sequences to identify novel AMPs.•Joker is capable of creating AMPs from inactive peptides.•Joker is capable of creating sequences with no similarity to known peptides.