RESUMO
Bacteriophages (phages) are viruses that infect bacteria. Many of them produce specific enzymes called depolymerases to break down external polysaccharide structures. Accurate annotation and domain identification of these depolymerases are challenging due to their inherent sequence diversity. Hence, we present DepoScope, a machine learning tool that combines a fine-tuned ESM-2 model with a convolutional neural network to identify depolymerase sequences and their enzymatic domains precisely. To accomplish this, we curated a dataset from the INPHARED phage genome database, created a polysaccharide-degrading domain database, and applied sequential filters to construct a high-quality dataset, which is subsequently used to train DepoScope. Our work is the first approach that combines sequence-level predictions with amino-acid-level predictions for accurate depolymerase detection and functional domain identification. In that way, we believe that DepoScope can greatly enhance our understanding of phage-host interactions at the level of depolymerases.
Assuntos
Bacteriófagos , Biologia Computacional , Bacteriófagos/genética , Bacteriófagos/enzimologia , Biologia Computacional/métodos , Anotação de Sequência Molecular , Proteínas Virais/genética , Proteínas Virais/metabolismo , Proteínas Virais/química , Redes Neurais de Computação , Aprendizado de Máquina , Software , Domínios Proteicos , Genoma Viral/genética , Hidrolases de Éster Carboxílico/genética , Hidrolases de Éster Carboxílico/metabolismo , Hidrolases de Éster Carboxílico/químicaRESUMO
IMPORTANCE: The emergence of multi-drug resistant bacteria is a global health problem. Among them, Klebsiella pneumoniae is considered a high-priority pathogen, making it necessary to develop new therapeutic tools to reduce the bacterial burden in an effective and sustainable manner. Phages, bacterial viruses, are very promising tools. However, phages are highy specific, rendering large-scale therapeutics costly to implement. This is especially certain in Klebsiella, a capsular bacterium in which phages have been shown to be capsular type dependent, infecting one or a few capsular types through specific enzymes called depolymerases. In this study, we have isolated and characterized novel phages with lytic ability against bacteria from a wide variety of capsular types, representing the Klebsiella phages with the widest range of infection described. Remarkably, these broad-range phages showed capsule dependency, despite the absence of depolymerases in their genomes, implying that infectivity could be governed by alternative mechanisms yet to be uncovered.