Análise metagenômica de rúmen bovino: revisão de algoritmos e processamento de dados para a compreensão do metagenoma ruminal

Abstract

Cattle farming plays a crucial role in food production, and understanding the rumen microbiota is essential for optimizing nutrition and productivity in this sector. The use of advanced Next-Generation Sequencing (NGS) techniques and metagenomic analysis enables the study of diversity and function within this complex microbial community. However, metagenomic analysis requires sophisticated algorithms and computational techniques to identify microorganisms, reconstruct genomes, and understand metabolic capabilities. Therefore, it is essential to be familiar with these algorithms to obtain meaningful results, address challenges, and improve analysis strategies. In this context, this study aimed to identify algorithms, process, and analyze metagenomic data from samples of bovine rumen subjected to natural additive supplementation. Data from a previous project available in the European Bioinformatics Institute database were also analyzed. Compatible algorithms were used with the infrastructure of a personal computer, including the Linux platform and Python distributions such as Anaconda 3 and Mamba. Steps such as data preprocessing, quality control, contig assembly, taxonomic identification, binning, and functional genome annotation were performed. The following algorithms were adopted for their respective analysis steps: FastQC, MultiQC, Trimmomatic, Megahit, Kraken 2, MetaBat2, MaxBin2, CONCOCT, Prokka, and SALMON. The results of sample processing demonstrated good quality, with the exception of one sample showing high duplication. Taxonomic analysis revealed a high percentage of unclassified reads but also identified differences in the abundance of bacterial genera such as Pseudoprevotella, Prevotella, and Alistipes. Specific species like Pseudoprevotella muciniphila were identified. The formation of Metagenome-Assembled Genomes (MAGs) highlighted the prevalence of certain taxonomic groups such as Bacteroidales and Clostridiales. Genome annotation identified hypothetical proteins and showed differences in genes related to lipases, cellulose metabolism, complex molecule breakdown, and antioxidants among the treatments. The study provides valuable information about the rumen microbiota, but it is important to note that future studies using updated tools and more detailed analyses offer promising opportunities to expand knowledge on the effectiveness of treatments in modulating rumen microbial composition.