Beierle, L. | bioRxiv Science

bioRxiv2026

First Genome-Scale Metabolic Model for Understanding HMPV-Host Interaction

BackgroundHuman Metapneumovirus (HMPV) is a major contributor to acute respiratory tract infections. Currently, no approved vaccines or specific antiviral therapies are available worldwide. Genome-scale metabolic models (GEMs), when integrated with Viral Biomass Objective Functions (VBOFs), provide a robust computational framework for identifying host metabolic dependencies essential for viral replication. This approach enables systematic prioritization of potential antiviral drug targets. ResultsThe first comprehensive VBOF for HMPV was constructed by integrating stoichiometric data from the viral genome, including structural proteins with defined copy numbers, amino acid residues per virion, envelope lipids, and glycan modifications. The reconstructed VBOF was incorporated into the human bronchial epithelial cell model, iHBEC1, to analyze metabolic changes between uninfected and infected host cells. Knockout analysis identified two selective antiviral gene targets: PGM3 (phosphoacetylglucosamine mutase) and GNPNAT1 (N-acetylglucosamine-6-phosphate acetyltransferase), as well as seven selective reaction targets primarily within the hexosamine biosynthesis and nucleotide sugar pathways. Knockout of PGM3 or GNPNAT1 completely abolished viral production while preserving complete host cell viability. Additionally, guanylate kinase (GUK1/GK1) emerged as a highly selective target for HMPV, confirming findings from Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) studies and suggesting a conserved vulnerability across respiratory viruses. ConclusionsUDP-GlcNAc, the vital end-product of the hexosamine biosynthesis pathway (HBP), is computationally predicted as a critical metabolic hub. This pathway represents the primary metabolic vulnerability of HMPV, due to the extensive glycosylation requirements of the HMPV attachment protein. The host-directed antiviral targets PGM3 and GNPNAT1 are high-priority candidates for experimental validation and may provide novel strategies to combat this respiratory infection.

Systems Biology↗

bioRxiv2025

Generative models for antimicrobial peptide design: auto-encoders and beyond

BackgroundSince the number of multi-resistant pathogens is growing rapidly, new strategies to accelerate the development of antimicrobial drugs are urgently needed. A promising candidate class for new antibiotics are antimicrobial peptides, showing lower tendency to induce antibiotic resistance. High-throughput in silico strategies for candidate mining, such as generative deep learning algorithms, have become popular over the last few years and offer novel ways for peptide discovery. MethodsThis study presents a comparative analysis of contemporary deep learning models generative performance for generating novel antimicrobial peptides. The models examined include Variational Auto-Encoders, a Wasserstein Auto-Encoder, a Recurrent Neural Network and a Language Model. The primary focus of this study is the systematic comparison and evaluation of various methods and sampling options to identify the most suitable model and sampling strategy combination for different use cases. ResultsThe findings demonstrate the models capacity to generate peptide sequences exhibiting analogous properties to those of naturally occurring active peptides, which are utilized for model training while featuring an appropriate degree of sequence diversity. Auto-encoder-based models, particularly the Wasserstein auto-encoder, have generated novel and remarkably diverse sequences compared to recurrent neural networks and language models. This model category exhibits a propensity to prioritize the frequencies of individual amino acids during the learning process, in contrast to variational auto-encoders. Furthermore, latent space models have been shown to possess the capacity to utilize diverse methodologies for generating novel peptides. However, it is imperative to note that these sampling strategies are not universally advantageous or disadvantageous; their optimal selection is contingent on the specificities of each individual use case. ConclusionThe present study investigates the strengths and weaknesses of various generative models for antimicrobial peptides and suggests which model and sampling strategy combination should be favoured for specific individual applications.

bioinformatics↗