Fig. 1

Reconstruction of the multi-kingdom microbial genomes of the goat gastrointestinal tract (GIT). a Sample collection along the goat GIT. A graphical representation of goat is shown with its GIT highlighted. The arrows along the GIT indicate the flow of food. The numbers in the parentheses next to the GIT site names indicate the samples obtained for this study. The GIT sites were divided into four sections in this study, namely stomach (rumen, reticulum, omasum, and abomasum), small intestine (duodenum, jejunum, and ileum), large intestine (cecum and colon) and rectum (fecal samples). The numbers beside the section names indicate the estimated numbers of microbes per nanogram DNA. b The rarefaction analysis of the unique number of bacterial/archaeal MAGs and viral genomes (Y-axis) as a function of sequenced samples (X-axis). The rarefaction curves for bacterial, archaeal, and viral genomes are shown in blue, green, and red, respectively (“Methods”). c Composition and quality of the genomes in the goat multi-kingdom microbiome catalog (GMMC), including 4004 bacterial and 71 archaeal MAGs and 7204 viral genomes. For GMMC MAGs, the quality criteria are defined by Bowers et al. [26]; complete: ≥ 90% completeness and ≤ 5% contamination according to CheckM [27] and at least 18 tRNA, high quality: ≥ 90% completeness and ≤ 5% contamination, medium quality: ≥ 80% completeness and ≤ 10% contamination, low quality: quality score (defined as the estimated completeness of a genome minus five times its estimated contamination) ≥ 50. For viral genomes, the quality is evaluated using CheckV [28]. d Contig N50 and N90 lengths (in bp) of GMMC genomes. e The mapping rates of clean reads to the GMMC genomes. The Wilcoxon rank sum test was used to show the statistical significance between groups; *** P < 0.001. f Percentages of novel bacterial and archaeal MAGs in GMMC as compared with public datasets at 95 and 99% average nucleotide identity (ANI) (“Methods”). g Mapping rates of metagenomic clean reads to the GMMC genomes as compared public datasets including the reference microbial genomes from the NCBI (BFAP, the combination of reference genomes including bacterial, fungal, archaeal, and protozoan reference genomes from the NCBI database) and MAGs of selected model organisms. The Wilcoxon rank sum test was used to show the statistical significance between groups; **** P < 0.0001)