Figure 1.
Schematic of the Atlantis II Deep brine pool and hypothetical bacterial niche-specific evolution.
The figure briefly describes the vertical changes of the temperature and salinity in the Atlantis II Deep brine pool. Brine water and a sediment core were collected from the deep layer of the brine pool and the seafloor, respectively. Microbial samples were then obtained at depths of 12 cm and 222 cm from the core. Free microbe exchanges occur at the interface of the seawater and brine pool, all under natural selection; the microbial lineages in the deep layer (a closed ecosystem) heavily evolve during their adaptation to the sediment.
Figure 2.
Microbial communities in the three samples.
The 16S fragments obtained from the reads were classified in the RDP database. The confidence threshold was not used in the classification. The genus with the highest confidence level was therefore assigned to the fragment. The average confidence level for genera is listed in Table S2. Only genera occupying >1% in one or more samples were included in the figure. All the remaining genera were included in minor groups.
Figure 3.
Number of reads for homologs per effective genome.
Three sets of read numbers for homologs used for the pairwise comparisons in which AIIBP, Sed12, and Sed222 were involved are shown in Figures 3a–3c, respectively. The homologs have more than 30 reads (longer than 300 bp) in both of the pairwise metagenomes. R values and equations of the correlations are also shown.
Figure 4.
Scores of z-tests using length of the longest ORFs in reads for homologs.
A z-test was performed to compare the lengths of the longest unbroken ORFs belonging to the homologs with >30 reads in both metagenomes. The z-scores for the homologs shared by three pairwise comparisons, AIIBP-Sed222, AIIBP-Sed12 and Sed12-Sed222, were plotted.
Figure 5.
Decay and persistence of abundant genes.
A total of 701 homologs (>0.05% of total reads in at least one sample) were used to show their abundance in three samples, AIIBP, Sed12 and Sed222. The genes were clustered by the Cluster 3 program using the complete linkage method after normalization of their abundances (a). The corresponding disruption rates of the homologs are shown in (b). The disrupted genes in the red part of (a) were removed according to the rates in (b). The heatmap (a) was thus modified to (c) showing the adjusted abundance of the homologs after the reduction of the disrupted genes.
Table 1.
Summary of the disruption rates of the abundant genes.
Table 2.
Statistics of the disruption styles.