João B. Xavier

The paper identifies resource-access breadth as a protective trait in non-toxigenic Clostridioides difficile strains, distinguishing protection from low disease potential through metabolic and competitive profiling.

The paper shows gut microbiota regulates the intestinal kynurenine pathway through vitamin E absorption, using mouse models and computational analysis of microbiota associations.

The paper identifies metabiosis, an ecological dependency where Bacteroides supports Pseudomonadota growth, increasing gut-borne bloodstream infection risk in allo-HCT patients.

This study identifies TDCA, a microbiota-derived bile acid, as a mediator that restores immune checkpoint therapy efficacy after antibiotic-induced hyporesponsiveness by enhancing CD8+ T cell antitumor immunity.

This study identifies an atypical EMT induced by TGF-β in lung adenocarcinoma metastatic progenitors that reduces biomechanical stiffness, enabling evasion of immune surveillance during dormancy.

The paper introduces pNODE, a model that predicts gut microbiome dynamics and bloodstream infections in cancer patients using clinical data and antibiotic perturbations.

The paper uses machine learning to identify valeric acid as a bacterial metabolite that inhibits Candida parapsilosis growth through intracellular acidification, reducing fungal colonization in mice.

This paper introduces MetaboLiteLearner, a machine learning framework that predicts metabolic changes in cells using GC/MS fragmentation patterns, demonstrating its ability to identify organ-specific metabolic adaptations in breast cancer metastasis.

The paper investigates the evolution and regulation of rhamnolipid biosynthesis in Pseudomonas aeruginosa, linking it to oxidative stress and explaining its inconsistent distribution across lineages.

The paper compiles shotgun metagenomic data from stool samples of hospitalized hematopoietic cell transplant patients to enhance microbiome analysis beyond taxonomic composition.

The paper shows that Pseudomonas aeruginosa uses integrated diffusive cues to stabilize cooperative swarming behavior against cheating.