AMPlify: attentive deep learning model for discovery of novel antimicrobial peptides effective against WHO priority pathogens

Abstract

Background Antibiotic resistance is a growing global health concern prompting researchers to seek alternatives to conventional antibiotics. Antimicrobial peptides (AMPs) are attracting attention again as therapeutic agents with promising utility in this domain, and using in silico methods to discover novel AMPs is a strategy that is gaining interest. Such methods can sift through large volumes of candidate sequences and reduce lab screening costs. Results Here we introduce AMPlify, an attentive deep learning model for AMP prediction, and demonstrate its utility in prioritizing peptide sequences derived from the Rana [Lithobates] catesbeiana (bullfrog) genome. We tested the bioactivity of our predicted peptides against a panel of bacterial species, including representatives from the World Health Organization’s priority pathogens list. Four of our novel AMPs were active against multiple species of bacteria, including a multi-drug resistant isolate of carbapenemase-producing Escherichia coli. Conclusions We demonstrate the utility of deep learning based tools like AMPlify in our fight against antibiotic resistance. We expect such tools to play a significant role in discovering novel candidates of peptide-based alternatives to classical antibiotics.

Identification and Functional Characterization of Peptides With Antimicrobial Activity From the Syphilis Spirochete, Treponema pallidum

“…In line with these results, few reports have been published describing this rare antimicrobial activity, making Tp0451a_C a new addition to this select group of S. pyogenes -targeting AMPs. Furthermore, the anti- S. pyogenes activity of Tp0451a_C was found to be generally comparable ( Cogen et al, 2010 ; Uhlmann et al, 2016 ; Ma et al, 2020 ) or more potent ( Sornwatana et al, 2018 ; Li et al, 2022 ) than known AMPs from other organisms. The relatively small number of AMPs that show activity against S. pyogenes is attributed to the expression of several proteins involved in AMP resistance, including the streptococcal cysteine protease SpeB ( Schmidtchen et al, 2002 ) which is involved in the proteolytic degradation and inactivation of AMPs, the M1 protein, streptokinase, and the streptococcal inhibitor of complement, the latter three of which are involved in resistance against defensins and/or LL-37 ( Frick et al, 2003 ; Lauth et al, 2009 ; Hollands et al, 2012 ).…”

“…It is also likely that T. pallidum encounters pathogenic / uropathogenic strains of E. coli during infection. Although these strains were not tested in our antimicrobial susceptibility assays, previous studies have shown that AMPs from other organisms exhibit similar levels of antimicrobial activity against both non-pathogenic and pathogenic / uropathogenic E. coli strains ( Fedders et al, 2010 ; Aghazadeh et al, 2019 ; Mardirossian et al, 2019 ; Moazzezy et al, 2020 ; Li et al, 2022 ; Lin et al, 2022 ). Tp0749_C also exhibited antimicrobial activity against the Gram-negative bacterium, P. aeruginosa .…”

“…For the experimental validation of antimicrobial and immunomodulatory activity, four putative AMPCCR peptides ( Table 1 ) from two T. pallidum miniproteins that were identified using our AMP bioinformatics prediction pipeline (Tp0451a_N, Tp0451a_C, Tp0749_N, and Tp0749_C), a cysteine-to-serine substituted version of Tp0749_C (Tp0749_C_C61S), and a cysteine-to-serine substituted version of Tp0451a_C (Tp0451a_C_C85S), were synthesized without chemical modifications using the PepPower™ solid state peptide synthesis (SSPS) platform at GenScript (NJ, United States). The known AMP, human cathelicidin LL-37 ( Turner et al, 1998 ), the known bullfrog ( Rana [Lithobates] catesbeiana ) AMP, RaCa-2 ( Li et al, 2022 ), a scrambled version of LL-37 (sLL-37), and a peptide (Tp0751_p5) from the T. pallidum adhesin Tp0751 ( Cameron et al, 2005 ) ( Table 1 ) were also synthesized via the same SSPS platform at GenScript, and used as positive (LL-37 and RaCa-2) and negative (sLL-37 and Tp0751_p5) controls in antimicrobial susceptibility and immunomodulation assays, as described below.…”

Improved Database Filtering Technology Enables More Efficient Ab Initio Design of Potent Peptides against Ebola Viruses

“…It is predicted that DDIP1 is effective against a variety of mutated viral strains, since it may damage viral envelopes to prevent infection in a similar manner to disrupt bacterial membranes. Our discovery of novel antiviral peptides may be further accelerated by applying the machine learning/artificial intelligence algorithms [ 35 , 36 , 37 , 38 , 39 , 40 , 41 , 42 , 43 ] that enable the prediction of both the antiviral activity and toxicity of peptides with the accumulation of experimental data of AMPs.…”

Krein Support Vector Machine Classification of Antimicrobial Peptides

“…Nonetheless, there has so far been a lack of success in translating AMP-based therapy to clinical use, due to challenges such as complex structure-activity relationship (SAR), drug toxicity, instability in host and infective environment, and low financial incentives 11,12 . Owing to the complex SAR and the costly and timeconsuming process of wet-lab experiments associated with AMP investigations, many researchers have proposed computational approaches, including molecular dynamics (MD) simulations and machine learning (ML) algorithms, to accelerate the discovery and development of potential AMPs for clinical use [13][14][15][16][17][18][19] .…”