DNA sequence data and bacterial community structure of NDF and ADF samples between high and low groups of Suhuai pigs
More than one million sequences were obtained from all samples, and there were 38,973 high quality sequences per sample and a range of 29,641 to 49,819. The average sequence length was 240bp. A total of 927 OTUs were identified from H-NDF and L-NDF groups, Core OTUs comprised approximately 94% of the total OTUs (Supplementary Figure 1A) in NAD group, while 45 and 11 OTUs were characteristically showed in H-NDF and L-NDF groups, respectively. At the same time, core OTUs comprised approximately 92% of the total OTUs (Supplementary Figure 1B) in ADF group, while 49 and 26 OTUs were characteristically showed in H-ADF and L-ADF groups, respectively.
Shannon and simpson indexes were significantly different between the two types of samples in NDF groups (P<0.05, Supplementary Table 2). There were no significantly different between the two types of samples in ADF groups (P>0.05, Supplementary Table 2). The microbiota of H-NDF and H-ADF clustered separately from the microbiota of the L-NDF and L-ADF along principal coordinate 1, respectively (Figure 1A, B). Microbial composition had a strong difference between high and low digestibility groups of NDF and ADF, respectively (Adonis/PERMANOVA, P<0.01, Supplementary Figure 2A, B).
Fourteen phyla were identifiedfrom the four groups (Supplementary Figure 3A): Firmicutes, Bacteroidetes, Actinobacteria, Tenericutes, Spirochaetae, Verrucomicrobia, Proteobacteria, Planctomycetes, unclassified_k_norank, Saccharibacteria, Cyanobacteria, Chlamydiae, Fibrobacteres and Lentisphaerae. Firmicutes and Bacteroidetes, comprised more than 91% of the total sequences, were the most predominant phyla in all samples. The abundances of Spirochaetae, Bacteroidetes and unclassified_k__norank were significantly different between H-NDF group and L-NDF group (P<0.05). In the meantime, the abundances of Spirochaetae, Verrucomicrobia, unclassified_k__norank and Fibrobactere were significantly different between H-ADF group and L-ADF group (P<0.05).
At genus level, 189 genera were identified from NDF samples, and 182 of those existing defined as core genera while 6 and 1 genera were uniquely identified in H- NDF and L- NDF, respectively (FigureS4 A). Meanwhile, 190 genera were distinguished from ADF samples, and 183 of those existing defined as core genera while 5 and 2 genera were uniquely identified in H-ADF and L-ADF, respectively (FigureS4 B). The 2 most dominant genera were Lactobacillus and Streptococcus belong to the phylum Firmicutes, comprised more than 25% and 47% of the total sequences in H-NDF group and L-NDF group, respectively (FigureS3 B). The 2 most predominant genera in H-ADF group and L-ADF group, separately containing about 44% and 35% of the total sequences, were Lachnospira and Streptococcus also belong to the phylum Firmicutes (FigureS3 D).
A total of 30 genera were found to be potential biomarkers between H-NDF group and L-NDF group; 29 genera were unique to H-NDF group and 1 WAS unique to L-NDF group (Figure3 A). At the same time, a total of 29 genera were found to be potential biomarkers between H-ADF group and L-ADF group; 23 genera were unique to H-ADF and 6 were unique to L-ADF (Figure3 B). Eubacterium coprostanoligenes group, Candidatus Soleaferrea, dgA 11 gut_group, Family XIII AD3011 group, norank f p 2534 18B5 gut group, norank f Porphyromonadaceae, Oscillibacter, Ruminococcaceae NK4A214 group, Succinivibrio, Treponema 2, unclassified f Ruminococcaceae and unclassified k norank were found to be potential biomarkersbetween H-NDF/H-ADF group and L-NDF/L-ADF group (H-NDF∩H-ADF, Figure3 A, B). Anaeroplasma, Anaerovibrio, Erysipelotrichaceae UCG 003, Lachnospiraceae ND3007 group, Lachnospiraceae NK4A136 group, Lachnospiraceae NK4B4 group, norank f Bacteroidales S24 7 group, norank f Mitochondria, Prevotella 1, Ruminiclostridium 1, Ruminococcaceae UCG 004, Ruminococcaceae UCG 005, Ruminococcus 1, Staphylococcus, unclassified f Lachnospiraceae, unclassified o Bacteroidales and unclassified p Bacteroidetes were found to be potential biomarkersonly between H-NDF group and L-NDF group (H-NDF, Figure3 A). Christensenellaceae R 7 group、Fibrobacter、norank c WCHB1 41、norank f Clostridiales vadinBB60 group、norank o Bradymonadales、Prevotella 2、Prevotellaceae UCG 001、Quinella、Ruminococcaceae UCG 002、Schwartzia and unclassified o Clostridiales were found to be potential biomarkers only between H-ADF group and L-ADF group (H-ADF, Figure3 B).
Prediction function of microbial metabolism
Sixty four functions were predicted in the present study. General function prediction only (8.56%), Carbohydrate transport and metabolism (8.25%), Amino acid transport and metabolism (8,23%), Replication (8.16%), Translation (7.85%), Transcription (7.50%), Cell wall/membrane/envelope biogenesis (6.47%), Energy production and conversion (5.51%), Inorganic ion transport and metabolism (5.08%) and Signal transduction mechanisms (4.49%)were the most enrichment functions. At the same time, 39 metabolic pathways were predicted. Membrane Transport (12.97%), Carbohydrate Metabolism (10.42%), Replication and Repair (9.69%), Amino Acid Metabolism (9.16%), Translation (6.47%), Energy Metabolism (5.54%), Poorly Characterized (4.81%), Nucleotide Metabolism (4.45%) and Metabolism of Cofactors and Vitamins (4.04%) were the most enrichment pathways.
According to the Clusters of Orthologous Groups of proteins (COG) and Kyoto Encyclopedia of Genes and Genomes (KEGG) databases, the top 10 in abundance predictive functions of potential biomarkers in H-NDF group and H-ADF group were shown in table 3, respectively. And the top 10 in abundance microbial metabolic pathways of potential biomarkers in H-NDF group and H-ADF group were shown in table 4, respectively. The most important functions and metabolic pathways of the above different potential biomarkers included carbohydrate transport and metabolism and Carbohydrate Metabolism, respectively.