登上Cell子刊封面：山东大学利用AI揭示发酵食品微生物组中的酶多样性

Core Viewpoint - The research highlights the hidden enzyme diversity and distribution within the microbiome of fermented foods, emphasizing the untapped potential for enzyme resource development in food research [2][3][8]. Group 1: Research Findings - The study utilized AI-assisted functional annotation to uncover enzyme diversity in the fermented food microbiome, providing valuable insights for future microbial function exploration in food research [3][10]. - The research team explored 10,202 metagenomic assembled genomes from global fermented foods, identifying over 5 million enzyme sequences categorized into 98,693 homologous clusters, representing more than 3,000 enzyme types [6]. - Functional analysis revealed that 84.4% of these clusters are unannotated in current databases, with terpenoid and polyketide metabolic enzymes showing high novelty [6]. Group 2: Environmental Adaptability - Peptidases exhibited broad environmental adaptability based on predicted optimal temperature and pH, with 31.3% of enzyme clusters demonstrating food type specificity [6]. - A machine learning model was developed to classify the source of fermented foods based on enzyme clusters, highlighting the potential for targeted optimization in food production [6]. Group 3: Related Commentary - A commentary article published in the same journal emphasizes that AI-assisted functional annotation reveals hidden microbial enzyme diversity and distribution, providing clues for elucidating ecological roles and biotechnological potential [9][10].