A bioinformatic pipeline and tailored workflow were developed to reconstruct Metagenome-Assembled Genomes (MAGs) from metagenomic data obtained from a cohort of 12 children with obesity and 12 normal-weight children with differential bisphenol A (BPA) exposure levels. Preprocessing and sequence quality control included FastQC, Trimmomatic and host DNA removal with Bowtie2. Assembly was performed using MEGAHIT, with quality evaluated via QUAST. Reads were mapped back to the assembly using Bowtie2 and SamTools to determine coverage. For Binning and Contig Clustering, MetaBAT2 and MaxBin2 were used, followed by optimization through the MetaWRAP bin_refinement module and quality validation (completeness/contamination) through CheckM. Tools for downstream analysis and MAG annotation included Prokka for Gene Prediction and Functional Gene Annotation, enabling the identification of key metabolic pathways. Taxonomic classification was performed using GTDB-Tk to provide standardized phylogenetic assignments. All statistical analyses and data visualization were conducted in RStudio.

Benchmarking of the tested tools revealed differences in the quality of the reconstructed MAGs across varying thresholds, with completeness ranging from >50% to >90% and contamination from <10% to <5%. A consolidated summary of tools and optimized parameters was compiled to evaluate pipeline performance based on completeness, contamination, N50, and taxonomic resolution. This comparative assessment enabled the identification of the most reliable assembly and binning strategies for downstream analyses.

The main results revealed distinct taxonomic signatures and quantitative differences between the two groups. MAG yield was lower in the obesity group compared to the normal-weight group. At the class level, obesity was associated with a distinct profile of Actinomycetes and Bacilli, whereas at the species level, a high prevalence of Bifidobacterium adolescentis was identified in children with obesity. Additionally, BPA exposure levels appeared to modulate microbial diversity and the relative abundance of specific taxa, suggesting a potential interaction between xenobiotic exposure and gut microbiome composition.