Skip to main content
NCBI home page
Bioinformatics and Biology Insights logoLink to Bioinformatics and Biology Insights
. 2023 Apr 14;17:11779322231166229. doi: 10.1177/11779322231166229

The Alteration of Akkermansiaceae/Lachnospiraceae Ratio Is a Microbial Feature of Antibiotic-Induced Microbiota Remodeling

Pei-Chen Chen 1,*, Ming-Shian Lin 2,*, Tien-Ching Lin 3, Ting-Wei Kang 1, Jhen-Wei Ruan 1,✉
PMCID: PMC10108413  PMID: 37077899

Abstract

Antibiotic treatment has been shown to cause gut microbiota dysbiosis. However, lacking critical features defining gut microbiota dysbiosis makes it challenging to prevent. By co-occurrence network analysis, we found that despite short antibiotic courses eliminating certain microbial taxa, the Akkermansia genus played the role of a high-centrality hub to maintain microbiota homeostasis. When the antibiotic courses continued, the elimination of Akkermansia induced a significant microbiota remodeling of the gut microbiota networks. Based on this finding, we found that under long-term antibiotic stress, the gut microbiota was rearranged into a stable network with a significantly lower Akkermansiaceae/Lachnospiraceae (A/L) ratio and no microbial hub. By functional prediction analysis, we confirmed that the gut microbiota with a low A/L ratio also had enhanced mobile elements and biofilm-formation functions that may be associated with antibiotic resistance. This study identified A/L ratio as an indicator of antibiotic-induced dysbiosis. This work reveals that besides the abundance of specific probiotics, the hierarchical structure also critically impacts the microbiome function. Co-occurrence analysis may help better monitor the microbiome dynamics than only comparing the differentially abundant bacteria between samples.

Keywords: Antibiotics, gut microbiota, dysbiosis, microbiota remodeling, network

Introduction

A mature ecosystem possesses high biodiversity and can maintain ecological stability and succession. The gut microbiota is a complex microbial community that can sustain a steady equilibrium against dynamic perturbations. 1 The ability to prevent severe pathogenic invasions, known as colonization resistance, is critical for healthy gut microbiota to maintain host homeostasis. 2 In addition, resilience is another essential capability for gut microbiota to tolerate pressure and maintain stability.1,3 However, gut microbiota compositions can be influenced by various factors, including diets, external environments, infections, and antibiotics. 4 Gut microbiota plasticity has been considered to play an important role in helping the host adapt to the challenges. 5 Moreover, changes in the composition of gut microbiota are also associated with many diseases and cancers. 6

Antibiotic treatment is an effective and efficient clinical therapy to eliminate life-threatening bacterial infections. However, antibiotics might cause negative effects on hosts, such as antibiotic-induced diarrhea and metabolic diseases.7,8 These symptoms and syndromes have been found to be strongly associated with alterations in the gut microbiota composition. 9 In healthy people, short-term antibiotic exposure could even lead to an imbalanced and disrupted gut microbiota community that requires a long time to reconstitute. 10 Long-term or repeated antibiotic treatment usually results in gut microbiota dysbiosis, which also impacts other distal organs. 11 Dysbiosis is defined as a hard-to-recover disturbance of gut microbial homeostasis. 12 Multiple microbial features might be observed in a dysbiotic microbiota, including low diversity, increased abundance of harmful microbes, and reduced abundance of beneficial microbes. 12 In healthy individuals, the well-structured gut microbiota with high microbial diversity can prevent colonization and growth of opportunistic pathogens, such as Clostridioides difficile. The reduced diversity and increased proportion of aerobic pathobionts in gut microbiota were often observed in C difficile colonized or infected patients. 13 However, no clear feature of gut microbiota changes has been suggested to be sufficient for concurrently monitoring dysbiosis during the antibiotic treatment period.

Akkermansia muciniphila is a gram-negative and strictly anaerobic bacterium belonging to the Verrucomicrobiae phylum. With the ability to degrade mucins, it has been detected in the mucosa layer near epithelial cells and might be associated with the mucosal-to-luminal regulation of microbiota composition.14,15 The abundance of A muciniphila in the gut has been demonstrated to negatively correlate with obesity and other metabolic disorders in humans and mice.16-18 Furthermore, supplementation with A muciniphila has shown beneficial effects on glucose tolerance and reducing fat mass in obese and diabetic mice and obese humans.19-22 Cold exposure has been shown to diminish the representation of A muciniphila by increasing the intestinal absorption surface in mice. 23 As one of the most abundant bacteria in the human gut, the regulation of A muciniphila abundance may play a critical role in regulating compositional homeostasis and the function of the gut microbiota.

Previous studies have proven the negative effects of antibiotics on the gut microbiota in low-diversity dysbiosis.9,10 However, the sequential changes of gut microbiota compositional succession remain unclear. In this study, we aimed to investigate how short-term and long-term antibiotic treatment courses influence gut microbiota using 16S rRNA sequencing. Our results revealed a significant shift in gut microbial diversity and composition during short-term and long-term antibiotic treatment. Co-occurrence network analysis sheds light on the loss of potential key microbes in short-term antibiotic treatment. Monitoring those candidates may be an early indicator to detect gut microbiota changes.

Methods

Sample collection

A total of 42 stool samples were collected longitudinally from 7 patients in intensive care unit (ICU) of National Cheng Kung University Hospital (NCKUH; IRB protocol number: A-ER-107-097). All patients were treated with broad-spectrum cepham antibiotics primarily due to urinary tract infection. The samples collected before the start of the antibiotic course were assigned to the untreated group. The samples collected at the time point less than 10 days during the antibiotic course were assigned to the short-term group. The samples collected at the time point of more than 10 days during the antibiotic course were assigned to the long-term group. The samples collected at the time point after the withdrawal of the antibiotic course were assigned to the recovery group. Fresh stool samples were preserved using DNA/RNA Shield-Fecal Collection Tube (Zymo Research, Irvine, CA, USA, Cat #R1101) and stored at −20°C before extraction.

Fecal DNA extraction

Fecal DNA was extracted from the mixture of stool and DNA/RNA shield (Zymo Research, Cat #R1100). The procedure was followed by the manufacturer’s illustration of ZymoBIOMICS DNA Miniprep Kit (Zymo Research, Cat #D4300). DNA concentration was detected using NanoDrop 2000/2000c Spectrophotometer (Thermo Scientific).

16S rRNA sequencing analysis

The V3-V4 region of 16S rRNA was amplified using the primer set of 341F and 805R that had been reported to have a low bias on amplifying V3-V4 amplicons. 24 A 300 bp length of paired-end 16S rRNA amplicons was performed on an Illumina MiSeq platform. 16S rRNA sequencing data were analyzed using QIIME 2 platform (version: 2021.4). Raw reads were demultiplexed by the q2-demux plugin, and quality control was performed using DADA2 25 via the q2-dada2 plugin that generated amplicon sequence variants (ASVs). Taxonomy was assigned to ASVs with classify-consensus-blast taxonomy classifier 26 through the q2-feature-classifier plugin using the SILVA database (release 138) with the 97% identity. Rarefaction was performed at 15 000 sequences per sample using R software. All samples were normalized to the minimum read count for differential sequencing depth among samples. Observed richness, evenness, phylogenetic diversity, and the Shannon diversity were calculated as indices of alpha diversity. For beta diversity, the weighted or unweighted-UniFrac distance and the Bray-Curtis distance were calculated within indicated groups using the vegan package (Community Ecology Package version 2.6-2) of R software. For the BugBase analysis, ASVs were clustered by the q2-vsearch plugin against Greengenes 13.8 database at a 97% identity threshold. For analysis of compositions of microbiomes with bias correction (ANCOM-BC) analysis, 27 the differential taxa were identified through the global test by false discovery rate (FDR) < 0.001. The global test showed differential taxa that were identified between at least 2 groups (Untreated vs Long-term or Short-term vs Long-term).

Co-occurrence network analysis

The SparCC 28 correlation coefficients were calculated between genera within 1 group. An edge was assigned if the correlation was higher than 0.6 with a P-value less than .05. Network features were estimated through degree centralization, betweenness centrality, and modularity using the igraph package of R software. 29 Betweenness centrality higher than 100 was colored by phylum. The network was visualized by Cytoscape software (version 3.9.1).

Statistical analysis

Statistical analysis, excluding ANCOM-BC, was performed on Prism 9 (GraphPad Software, CA, USA). Comparisons between 2 groups were analyzed using a Mann-Whitney test. Comparisons of more than 2 groups were performed by one-way analysis of variance (ANOVA) followed by the Tukey post hoc test. Differences were considered significant when P-value was smaller than .05.

Data availability

The 16S rRNA sequencing data have been deposited to National Center for Biotechnology Information (NCBI) under the following BioProject ID: PRJNA929331. This study does not report the original code. Any additional information required to reanalyze the data reported in this article is available from the lead contact on request.

Results

The time course of antibiotic treatment differentially alters gut microbiota composition

Stool samples were collected from patients with urinary tract infections in the ICU of NCKUH. According to the collected antibiotic (cephems) treatment time points, 42 stool samples were divided into 4 groups, including the untreated group (samples collected before treatments, N = 11), short-term group (samples collected during the treatment period shorter than 10 days, N = 16), long-term group (samples collected during the treatment period longer than 10 days, N = 11), and recovery group (samples collected after the termination of a short-term antibiotic course, N = 4). By 16S rDNA sequencing, we found that both short-term and long-term antibiotic treatment showed effects on changing the compositions of the gut microbiota (Figure 1A and B and Tables 1 and 2). The most dramatic changes were observed in the long-term treated microbiota that has increased proportions of Enterococcaceae and Lachnospiraceae families (Figure 1B and Table 2). By measuring alpha diversity, we found that long-term exposure to cephems antibiotic significantly reduced the observed richness, evenness, phylogenetic diversity, and the Shannon diversity when compared with those of the untreated group, but samples in the short-term treatment group did not show significant changes in these 4 metrics of alpha diversity compared with the untreated group (Figure 1C to F). Despite this, we still observed that the abundance of certain species was reduced by short-term antibiotic treatment, including Lactobacillus salivarius, Ruthenibacterium lactatiformans, Lachnoclostridium urinimassiliense, Bifidobacterium pseudocatenulatum, Streptococcus anginosus, and Bacteroides kribbi (Figure 1G). This suggests that these reduced species during the short-term treatment stage might be more sensitive to cephams than other gut commensals. These results suggest that the gut microbiota has adaptive plasticity for the short period of antibiotic exposure, but long-term treatments of antibiotic treatments may cause a significant reduction in microbial diversity.

Figure 1.

Figure 1.

Open in a new tab

The time course of antibiotic treatment differentially alters gut microbiota composition: (A–B) composition and (C–F) alpha diversity of gut microbiota in untreated, short-term, long-term, and recovery groups. Bar plots showed microbial relative abundances at the phylum level (A) and at the family level (B). Alpha diversities were determined using (C) observed richness, (D) evenness, (E) phylogenetic diversity, and (F) the Shannon diversity. (G) The eliminated bacteria in short-term antibiotic treatment, including Lactobacillus salivarius, Ruthenibacterium lactatiformans, Lachnoclostridium urinimassiliense, Bifidobacterium pseudocatenulatum, Streptococcus anginosus, and Bacteroides kribbi. Data are presented by mean ± SEM. *P < .05; **P < .01; ***P < .001; ****P < .0001 according to the one-way ANOVA analysis and the Tukey post hoc test (C–F) and Mann-Whitney analysis (G). Untreated, N = 11; Short-term, N = 16; Recovery, N = 4. ns indicates non-significant; ANOVA, analysis of variance.

Table 1.

Microbiota composition at phylum level (related to Figure 1A).

Untreated Short-term Long-term Recovery Untreated vs short-term Short-term vs long-term Untreated vs long-term Untreated vs recovery Short-term vs recovery Long-term vs recovery
Actinobacteriota 2.19 ± 0.42% 1.71 ± 0.45% 4.43 ± 1.08% 1.06 ± 0.08% P = .9479 P = .0200* P = .1123 P = .8277 P = .9556 P = .0687
Bacteroidota 27.54 ± 1.42% 24.92 ± 2.16% 14.25 ± 1.85% 23.61 ± 2.01% P = .7592 P = .0016** P = .0003*** P = .7570 P = .9860 P = .1042
Firmicutes 44.49 ± 1.8% 45.46 ± 2.94% 64.42 ± 6.48% 55.41 ± 2.33% P = .9978 P = .0156* P = .0196* P = .5273 P = .5685 P = .8169
Fusobacteriota 2.03 ± 0.45% 3.42 ± 0.92% 0.5 ± 0.14% 1.64 ± 0.71% P = .4889 P = .0231* P = .4780 P = .9928 P = .5767 P = .8609
Proteobacteria 20.18 ± 2.69% 22.36 ± 2.83% 17.94 ± 5.72% 14.95 ± 1.38% P = .9729 P = .8181 P = .9768 P = .8987 P = .7350 P = .9786
Synergistota 1.61 ± 0.47% 0.68 ± 0.19% 0.1 ± 0.03% 2.49 ± 0.16% P = .0719 P = .3986 P = .0029** P = .3855 P = .0073** P = .0005***
Open in a new tab

Data are presented by mean ± SEM. *P < .05, **P  < .01, ***P  < .001 according to One-Way ANOVA analysis and the Tukey post-hoc test.

Table 2.

Microbiota composition at family level (related to Figure 1B).

Untreated Short-term Long-term Recovery Untreated vs short-term Short-term vs long-term Untreated vs long-term Untreated vs recovery Short-term vs recovery Long-term vs recovery
Bacteroidaceae 19.62 ± 2.15% 19.21 ± 2.01% 11.27 ± 1.75% 18.48 ± 1.99% P = .9988 P = .0303* P = .0385* P = .9923 P = .9977 P = .3057
Bifidobacteriaceae 0.97 ± 0.26% 0.49 ± 0.13% 4.11 ± 1.12% 0.52 ± 0.14% P = .9271 P = .0002*** P = .0036** P = .9804 P > .9999 P = .0189*
Coriobacteriaceae 0.83 ± 0.37% 0.97 ± 0.44% 0.17 ± 0.09% 0.42 ± 0.21% P = .9924 P = .4088 P = .6400 P = .9481 P = .8714 P = .9882
Enterobacteriaceae 17.62 ± 2.98% 16.94 ± 2.52% 17.04 ± 5.73% 11.19 ± 2.39% P = .9991 P > .9999 P = .9996 P = .8235 P = .8510 P = .8606
Enterococcaceae 3.9 ± 1.36% 10.43 ± 3.54% 24.66 ± 7.27% 0.78 ± 0.43% P = .7025 P = .1032 P = .0157* P = .9856 P = .6800 P = .0540
Fusobacteriaceae 2.03 ± 0.45% 3.42 ± 0.92% 0.5 ± 0.14% 1.64 ± 0.71% P = .4902 P = .0233* P = .4787 P = .9929 P = .5782 P = .8607
Lachnospiraceae 18.76 ± 1.43% 11.9 ± 1.4% 20.01 ± 2.42% 17.1 ± 3.2% P = .0363* P = .0099** P = .9651 P = .9674 P = .4484 P = .8527
Lactobacillaceae 3.89 ± 0.23% 3.1 ± 0.87% 6.85 ± 1.84% 4.52 ± 0.81% P = .9529 P = .0796 P = .2911 P = .9922 P = .9115 P = .7324
Marinifilaceae 1.1 ± 0.24% 0.85 ± 0.16% 0.1 ± 0.05% 1.56 ± 0.23% P = .7105 P = .0114* P = .0016** P = .5431 P = .1561 P = .0007***
Oscillospiraceae 1.59 ± 0.35% 2 ± 0.5% 0.24 ± 0.11% 3.19 ± 0.41% P = .8784 P = .0149* P = .1320 P = .2308 P = .4477 P = .0053**
Prevotellaceae 2.24 ± 0.62% 1.93 ± 0.38% 1.01 ± 0.33% 1.45 ± 0.21% P = .9541 P = .4285 P = .2517 P = .8141 P = .9438 P = .9599
Rikenellaceae 1.34 ± 0.34% 0.72 ± 0.28% 1.02 ± 0.5% 0.43 ± 0.08% P = .5842 P = .9271 P = .9291 P = .6021 P = .9767 P = .8529
Ruminococcaceae 7.89 ± 0.98% 9.69 ± 1.79% 5.23 ± 1.09% 11.05 ± 2.61% P = .8247 P = .1625 P = .6527 P = .7424 P = .9681 P = .2604
Streptococcaceae 2.61 ± 0.53% 3.83 ± 1.04% 2.1 ± 0.88% 9.15 ± 2.92% P = .8147 P = .6023 P = .9868 P = .0158* P = .0509 P = .0083**
Sutterellaceae 2.2 ± 0.64% 5.26 ± 2.58% 0.86 ± 0.32% 3.29 ± 2.02% P = .6519 P = .3483 P = .9647 P = .9925 P = .9519 P = .9245
Synergistaceae 1.61 ± 0.47% 0.68 ± 0.19% 0.1 ± 0.03% 2.49 ± 0.16% P = .0719 P = .3986 P = .0029** P = .3855 P = .0073** P = .0005***
Tannerellaceae 2.61 ± 1.04% 1.79 ± 0.42% 0.56 ± 0.13% 1.39 ± 0.3% P = .7474 P = .4382 P = .1123 P = .7477 P = .9857 P = .9016
Open in a new tab

Data are presented by mean ± SEM. *P < .05, **P  < .01, ***P  < .001 according to One-Way ANOVA analysis and the Tukey post-hoc test.

By unweighted- and weighted-UniFrac analyses, we confirmed that the long-term antibiotic-treated microbiota were significantly different from the gut microbiota in the untreated, short-term treated, and recovery groups (Figure 2A and B). To understand the microbial succession during antibiotic exposure, we compared the gut microbiota of the short-term or long-term group with that of the untreated group by dissimilarity analysis of unweighted or weighted-UniFrac distances. Short-term treated microbiota and long-term treated microbiota had an increased dissimilarity based on unweighted-UniFrac distances compared with untreated microbiota (Figure 2C). This suggests that both short-term and long-term treatments induced significant presence-absence changes in the microbiota. Interestingly, only long-term but not short-term treated microbiota showed a significantly higher dissimilarity compared with untreated microbiota based on weighted-UniFrac distances (Figure 2D). This implies that a critical remodeling of gut microbiota may occur between the short-term/long-term transition of antibiotic exposure.

Figure 2.

Figure 2.

Open in a new tab

The ecological succession of antibiotic-treated microbiota. (A–B) Principal co-ordinates analysis (PCoA) of unweighted (A) and weighted (B) UniFrac distances of gut microbiota community in untreated, short-term, long-term, and recovery groups. Each dot represents a sample. Statistical analysis was performed using PERMANOVA pairwise test. (C–D) Dissimilarity analysis of unweighted (C) and weighted (D) UniFrac distances between gut microbiota of the untreated group and other indicated groups. (E) Heatmap of differentially abundant genera identified by ANCOM-BC analysis. Data are presented by mean ± SEM. *P < .05; **P < .01; ***P < .001; ****P < .0001 according to the one-way ANOVA analysis and the Tukey post hoc test. Untreated, N = 11; short-term, N = 16; long-term, N = 11; recovery, N = 4. ANCOM-BC indicates analysis of compositions of microbiomes with bias correction; ANOVA, analysis of variance; ns, non-significant; PERMANOVA, permutational multivariate analysis of variance.

Long-term antibiotic treatment remodels the microbial network

By ANCOM-BC global test, we confirmed that a global change of multiple genera occurred specifically in long-term treated microbiota (Figure 2E). When compared with the untreated and short-term groups, long-term antibiotic exposure significantly reduced the abundance of gut bacteria in Clostridia, Bacilli, and Verrucomicrobiae classes (Figure 2E). To further understand whether the long-term antibiotic treatments induce the network remodeling of the gut microbiota, we performed SparCC co-occurrence network analyses to determine the interactive complexity of bacteria in the microbiota by the number of nodes and edges. 30 We found that although the number of nodes (genus-level taxa) was similar between untreated microbiota (node number = 98) and short-term treated microbiota (node number = 95), the edge number of the short-term treated microbiota was decreased from 399 to 168 (Figure 3A and B). Higher modularity was also observed in the short-term treated network, suggesting that the antibiotic treatment disrupted the compact network connections observed in untreated microbiota (Figure 3B). Intriguingly, despite that the long-term treatments dramatically reduced the nodes (nodes number = 41) in the microbial network, an increased number of correlated edges compared with short-term treated microbiota was observed (edge number = 211) (Figure 3C). This indicates that the prolonged antibiotic treatment can remodel the gut microbiota into a newly formed network that rebuilt the interactions between microbes (Figure 3C).

Figure 3.

Figure 3.

Open in a new tab

Long-term antibiotic treatment remodels the microbial network. (A–C) Co-occurrence networks are performed on SparCC correlations at the genus level of the (A) untreated microbiota, (B) short-term treated microbiota, and (C) long-term treated microbiota. Solid and dashed lines represent the positive correlation and negative correlation, respectively. Node sizes are characterized by betweenness centrality and betweenness centralities that are higher than 100 are colored by phylum. (D–F) Centralization measurement of degree and betweenness in the genus-level co-occurrence networks of the (D) untreated microbiota, (E) short-term treated microbiota, and (F) long-term treated microbiota. (G–H) Betweenness centralization (G) and degree centralization (H) in untreated, short-term treated, and long-term treated microbiota networks. Data are presented by mean ± SEM. *P < .05; **P < .01; ***P < .001; ****P < .0001 according to the one-way ANOVA analysis and the Tukey post hoc test. ANOVA indicates analysis of variance; ns, non-significant.

The microbial hub theory suggests that specific species in the human gut microbiota may play a role as a regulatory hub in maintaining homeostasis.31,32 Degree centralities and betweenness centralities are often used measurements for exploring the taxa with the hub-like property in a microbial network. Hence, we determined the microbial hubs in each network by setting a cutoff of betweenness centralities of nodes higher than 200 at the genus level. Before antibiotic treatment, the untreated network harbored multiple high-betweenness taxa with high degree centralities (Figure 3D). After the short-term treatment, most microbial hubs lost their betweenness and degree centralities to coordinate the network (Figure 3E). Instead, Akkermansia became the high-betweenness hub with the highest degree centrality in the network (Figure 3E). This suggests that Akkermansia plays a critical role in maintaining community stability during the antibiotic treatment period. However, in the long-term treated microbiota, no high-betweenness taxa were observed, suggesting that the hierarchical community structure was dramatically changed into a network with no regulatory hub (Figure 3F and G). Different from short-term treated microbiota, multiple genera in the Clostridia class showed high degree centralities in the network in long-term treated microbiota, including Lachnospira, Tyzzerella, and Ruminococcus gauvreauii groups (Figure 3F). Notably, the long-term treated microbiota dramatically rebuilt connections between microbes that significantly restore the degree centrality of the network (Figure 3H). This indicates that long-term antibiotic stress can remodel gut microbiota into a reorganized, stable community structure. The lack of microbial hub in the long-term treated network may also explain why dysbiotic microbiota is challenging to recover.

Akkermansiaceae/Lachnospiraceae ratio is an important feature of antibiotic-induced microbiota remodeling

Because the hierarchies of Akkermansia and Lachnospiraceae in microbiota networks were altered between the transition of short-term and long-term antibiotic treatments (Figure 3E and F), we further confirmed the changes of the abundance of Akkermansia and Lachnospiraceae in the short-term and long-term groups. Our results confirmed that Akkermansia was significantly eliminated in long-term treated microbiota, accompanied by a significant increase of Lachnospiraceae when compared with short-term treated microbiota (Figure 4A and B). To understand whether the features of Akkermansia-induced remodeling could be used for monitoring gut microbiota dysbiosis, we calculated ratios between Verrucomicrobiota and other phyla at different stages of the antibiotic course. At the phylum level, both Verrucomicrobiota/Firmicutes ratio and Verrucomicrobiota/Proteobacteria ratios were found to be significantly decreased in the long-term group compared with the untreated group (Table 3). At the class level, Verrucomicrobiae/Clostridia (V/C) and Verrucomicrobiae/Gammaproteobacteria ratios reflected the same trends (Table 3). At the family level, Akkermansiaceae/Lachnospiraceae (A/L) ratio and Akkermansiaceae/Enterobacteriaceae ratio were consistent with the results on phylum and class levels (Table 3). Based on the results of ANCOM-BC analysis, nearly half (37 out of 81) affected genera (long-term vs short-term) by long-term antibiotic treatment belong to the Clostridia class, and only 6 genera in the Gammaproteobacteria phylum were affected by long-term antibiotic treatment (Figure 2E). Therefore, these findings suggest that V/C and A/L ratios may be more sensitive than Verrucomicrobiota/Proteobacteria and Verrucomicrobiae/Gammaproteobacteria ratios to identify the occurrence of microbiota remodeling. Both A/L and V/C ratios were significantly decreased by long-term antibiotic exposure (Figure 4C and D and Table 3). These results suggest A/L ratio is an important feature of antibiotic-induced microbiota remodeling.

Figure 4.

Figure 4.

Open in a new tab

Akkermansiaceae/Lachnospiraceae ratio is an important feature of antibiotic-induced microbiota remodeling. (A–B) Relative abundances of the Akkermansiaceae family (A) and the Lachnospiraceae family (B) in short-term and long-term antibiotic treatment groups. (C–D) The Akkermansiaceae/Lachnospiraceae (A/L) ratio (C) and the Verrucomicrobiae/Clostridia (V/C) ratio (D) were calculated to perform the dysbiotic remodeling during antibiotic treatment. (E–H) Predicted phenotypes by the BugBase analysis, including aerobic (E), anaerobic (F), containing mobile elements (G), and biofilm formation (H). Data are presented by mean ± SEM. *P < .05; **P < .01; ***P < .001; ****P < .0001 according to the one-way ANOVA analysis and the Tukey post hoc test. Untreated, N = 11; short-term, N = 16; long-term, N = 11; recovery, N = 4. ANOVA indicates analysis of variance; ns, non-significant.

Table 3.

The Verrucomicrobiota-related changes at phylum, class, and family levels..

Mean ± SEM Untreated Short-term Long-term Recovery Untreated vs short-term Untreated vs long-term Untreated vs recovery
Phylum
 Verrucomicrobiota/Actinobacteriota 0.6399 ± 0.2248 0.5122 ± 0.1598 0.0516 ± 0.0225 0.1441 ± 0.1441 P = .8948 P = .0501 P = .3216
 Verrucomicrobiota/Bacteroidota 0.049 ± 0.0192 0.0363 ± 0.0157 0.0054 ± 0.0017 0.0064 ± 0.0064 P = .8697 P = .1379 P = .3726
 Verrucomicrobiota/Firmicutes 0.028 ± 0.0101 0.0176 ± 0.0059 0.0016 ± 0.0007 0.0037 ± 0.0037 P = .5337 P = .0280* P = .1881
 Verrucomicrobiota/Fusobacteriota 0.7859 ± 0.2651 0.6078 ± 0.2816 0.1473 ± 0.0364 0.0489 ± 0.0489 P = .9126 P = .2036 P = .3282
 Verrucomicrobiota/Proteobacteria 0.0753 ± 0.0278 0.0414 ± 0.0148 0.0065 ± 0.0018 0.0115 ± 0.0115 P = .3655 P = .0309* P = .1943
Class
 Verrucomicrobiae/Clostridia 0.0265 ± 0.009 0.0021 ± 0.0006 0.0046 ± 0.0046 0.0387 ± 0.0146 P = .6898 P = .0410* P = .2216
 Verrucomicrobiae/Gammaproteobacteria 0.0389 ± 0.0147 0.0057 ± 0.0016 0.0106 ± 0.0106 0.0742 ± 0.0277 P = .3334 P = .0310* P = .1949
Family
 Akkermansiaceae/Lachnospiraceae 0.0532 ± 0.0158 0.0031 ± 0.0007 0.0065 ± 0.0065 0.656 ± 0.0254 P = .9137 P = .0459* P = .2260
 Akkermansiaceae/Enterobacteriaceae 0.0508 ± 0.0173 0.0059 ± 0.0017 0.0265 ± 0.0265 0.0981 ± 0.0379 P = .3152 P = .0263* P = .3033
Open in a new tab

Data are presented by mean ± SEM. *P < .05 according to One-Way ANOVA analysis and the Dunnett’s multiple comparisons test.

To understand whether the long-term antibiotic-induced remodeling changes microbiota functions, functional prediction analysis was performed using the BugBase database. The results showed that the proportions of aerobic bacteria and anaerobic bacteria were both changed specifically by the long-term treatment but not by the short-term treatment (Figure 4E and F). This result supports our finding that the microbiota remodeling occurred specifically in long-term antibiotic-treated patients.

In addition, the long-term treated microbiota harbored a significantly higher proportion of bacteria mobile elements and biofilm-formation functions that may increase the pathogenic and antibiotic resistance of the gut microbiota (Figure 4G and H). By the BugBase analyses, the increase in biofilm-forming population was primarily contributed by biofilm-related taxa in Proteobacteria phylum (untreated: 8.665 ± 2.149%; long-term: 26.847 ± 6.741%) and biofilm-related taxa in Actinobacteria phylum (untreated: 3.256 ± 1.118%; long-term: 6.228 ± 2.684%) (Figure 4H). These results suggest that long-term antibiotic-induced microbiota remodeling could promote biofilm formation and drive the transfer of mobile elements that contribute to the stabilization of newly formed microbial social networks.33,34

Discussion

Although antibiotic treatment may induce gut microbiota dysbiosis, it is an essential treatment to prevent or eradicate bacterial pathogenic infections. To date, there is no sufficient laboratory diagnostic method for the identification of gut dysbiosis. In this study, we monitored the effect of short-term and long-term courses of antibiotics on human gut microbiota. We found that although the short-term antibiotic course may temporally collapse the microbial network, most microbial changes that occurred in this stage were irreversible. However, the long-term treatment course could induce microbiota remodeling to construct a rearranged microbial network with enhanced biofilm-formation function. The A/L ratio identified in this study may serve as indicator of microbiota dysbiosis. Accordingly, we offer practical indicators to help physicians be aware of the occurrence of gut microbiota dysbiosis synchronously during the treatment period.

Dysbiosis is a difficult-to-recover state of the gut microbiota with an illness phenotype and reflects an imbalanced interaction between hosts and microbes. 1 The elimination of certain gut bacteria may not simply induce the occurrence of dysbiosis because a healthy gut microbiota usually possesses a strong resilience to environmental stresses. 1 In this study, we found that besides the A/L ratio, the long-term antibiotic-treated microbiota also has significant alterations in the proportion of aerobes and anaerobes. Previous studies have demonstrated that oxygen influx is one of the critical factors needed to change microbial metabolic interactions, which can dynamically regulate the switching process.35,36 Because incomplete modulation of oxygen influxes may irreversibly impact the gut microbiota dysbiosis, it is suggested that recording the intestinal oxygen influx or the anaerobe ratio might be an alternative method to track the intestinal condition. Further investigation will be required to understand whether the A/L dynamics are associated with intestinal oxygen regulation.

Notably, we observed a substantial increase in Firmicutes and Actinobacteria in comparison to others in the long-term antibiotic-treated patient. Similar changes were also observed in previous studies investigating the effects of broad-spectrum antibiotics on gut microbiota in humans and gorillas.37,38 The potential explanation is that only a limited number of commensals have been found to colonize or adherent to the intestinal mucus layer, including Lachnospiraceae, Ruminococcaceae, Bifidobacterium bifidum, Bifidobacterium longum, and A muciniphila. 15 When prolonged antibiotic exposure eliminates the representation of A muciniphila, the diminishment of this dominant mucus-degrading species may subsequently support the expansion of other mucus-associated commensals, such as Lachnospiraceae (Firmicutes) and Bifidobacterium (Actinobacteria) identified in this study. This strengthens that A muciniphila is a critical microbial hub regulating gut microbiota remodeling.

Although no significant compositional change on phylum level was observed in short-term antibiotic-treated microbiota, the abundance of certain bacteria was specifically reduced by short-term exposure, such as B pseudocatenulatum, R lactatiformans, and B kribbi. All these 3 species are anaerobes that can produce lactic acid by glucose fermentation.39,40B pseudocatenulatum has been shown to alter the gut microbiota composition to protect mice against colitis and liver injury.41,42R lactatiformans has been shown to regulate intestinal immune responses.43,44 Two of the most significantly reduced species in short-term treated microbiota, B pseudocatenulatum and L salivarius, are strong probiotics that may be the chief cogs of the following A muciniphila-induced remodeling. It is worth considering that losing these species at the early stage of the antibiotic course may impact long-term microbiota succession by interfering host-microbe interaction when the treatment continues.

Despite Faecalibacterium, Akkermansia, and Lactobacillus genera showing higher degree centralities than other genera in long-term treated microbiota, the betweenness centrality of gut microbiota was dramatically interrupted by the prolonged antibiotic treatment. These findings suggest that the microbial hub, such as Akkermansia, is critical for maintaining the hierarchical structure of gut microbiota. Antibiotics globally impact the gut microbiota, including pathobionts, commensals, and probiotics.45,46 Decreased abundances of beneficial strains in the gut microbiota have been shown to be associated with intestinal permeability, inflammation, and metabolic disorders. 47 However, supplementation with probiotics after antibiotic treatment may interfere with the natural recovery of a healthy gut microbiome. 48 Hence, it suggests that the maintenance of a homeostatic gut microbiota may not simply depend on the presence or absence of specific probiotics. Carefully monitoring the dysbiosis-associated features of the gut microbiota in patients receiving long-term antibiotic treatment would be necessary for preventing the occurrence of dysbiosis and severe infections, such as C difficile infection.

Taken together, our results demonstrate that long-term antibiotic exposure induces substantial stress that causes remodeling and changing of the ecological functions of the gut microbiota. It is indispensable to identify the specific features of an imbalanced gut microbiota, which could prevent the occurrence of a difficult-to-recover dysbiosis.

Acknowledgments

The authors thank National Center for High-performance Computing (NCHC) for providing computational and storage resources in bioinformatics analyses.

Footnotes

Funding: The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The authors thank Ministry of Science and Technology for providing grant supports: 107-2320-B-006-020-MY2 and 108-2320-B-006-051-MY3. The authors also thank National Cheng Kung University Hospital for providing grant supports: NCKUH-10802034 and NCKUH-10902062. The authors also thank CHIA-YI CHRISTIAN HOSPITAL for providing grant supports: CYC110003.

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Author Contributions: T-CL and J-WR conceptualized this study. T-CL contributed to clinical sample acquisition and collection. T-WK isolated stool DNA. P-CC analyzed the sequencing data and wrote the original manuscript. T-CL, M-SL, and J-WR provided the funding to support this study. J-WR supervised the findings in the work. All authors reviewed the manuscript.

References

  • 1.Sommer F, Anderson JM, Bharti R, Raes J, Rosenstiel P. The resilience of the intestinal microbiota influences health and disease. Nat Rev Microbiol. 2017;15:630-638. doi: 10.1038/nrmicro.2017.58. [DOI] [PubMed] [Google Scholar]
  • 2.Pickard JM, Núñez G. Pathogen colonization resistance in the gut and its manipulation for improved health. Am J Pathol. 2019;189:1300-1310. doi: 10.1016/j.ajpath.2019.03.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Fassarella M, Blaak EE, Penders J, Nauta A, Smidt H, Zoetendal EG. Gut microbiome stability and resilience: elucidating the response to perturbations in order to modulate gut health. Gut. 2021;70:595-605. doi: 10.1136/gutjnl-2020-321747. [DOI] [PubMed] [Google Scholar]
  • 4.Sommer F, Bäckhed F. The gut microbiota—masters of host development and physiology. Nat Rev Microbiol. 2013;11:227-238. doi: 10.1038/nrmicro2974. [DOI] [PubMed] [Google Scholar]
  • 5.Alberdi A, Aizpurua O, Bohmann K, Zepeda-Mendoza ML, Gilbert MTP. Do vertebrate gut metagenomes confer rapid ecological adaptation. Trends Ecol Evol. 2016;31:689-699. doi: 10.1016/j.tree.2016.06.008. [DOI] [PubMed] [Google Scholar]
  • 6.Fan Y, Pedersen O. Gut microbiota in human metabolic health and disease. Nat Rev Microbiol. 2021;19:55-71. doi: 10.1038/s41579-020-0433-9. [DOI] [PubMed] [Google Scholar]
  • 7.Li J, Yang K, Ju T, et al. Early life antibiotic exposure affects pancreatic islet development and metabolic regulation. Sci Rep. 2017;7:41778. doi: 10.1038/srep41778. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Bartlett JG. Clinical practice. Antibiotic-associated diarrhea. N Engl J Med. 2002;346:334-339. doi: 10.1056/NEJMcp011603. [DOI] [PubMed] [Google Scholar]
  • 9.Ramirez J, Guarner F, Bustos Fernandez L, Maruy A, Sdepanian VL, Cohen H. Antibiotics as major disruptors of gut microbiota. Front Cell Infect Microbiol. 2020;10:572912. doi: 10.3389/fcimb.2020.572912. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Palleja A, Mikkelsen KH, Forslund SK, et al. Recovery of gut microbiota of healthy adults following antibiotic exposure. Nat Microbiol. 2018;3:1255-1265. doi: 10.1038/s41564-018-0257-9. [DOI] [PubMed] [Google Scholar]
  • 11.Worby CJ, Schreiber HL, IV, Straub TJ, et al. Longitudinal multi-omics analyses link gut microbiome dysbiosis with recurrent urinary tract infections in women. Nat Microbiol. 2022;7:630-639. doi: 10.1038/s41564-022-01107-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Wei S, Bahl MI, Baunwall SMD, Hvas CL, Licht TR. Determining gut microbial dysbiosis: a review of applied indexes for assessment of intestinal microbiota imbalances. Appl Environ Microbiol. 2021;87. doi: 10.1128/AEM.00395-21. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Seekatz AM, Young VB. Clostridium difficile and the microbiota. J Clin Invest. 2014;124:4182-4189. doi: 10.1172/JCI72336. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Derrien M, Van Baarlen P, Hooiveld G, et al. Modulation of mucosal immune response, tolerance, and proliferation in mice colonized by the mucin-degrader Akkermansia muciniphila. Front Microbiol. 2011;2:166. doi: 10.3389/fmicb.2011.00166. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Paone P, Cani PD. Mucus barrier, mucins and gut microbiota: the expected slimy partners. Gut. 2020;69:2232-2243. doi: 10.1136/gutjnl-2020-322260. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Everard A, Belzer C, Geurts L, et al. Cross-talk between Akkermansia muciniphila and intestinal epithelium controls diet-induced obesity. Proc Natl Acad Sci USA. 2013;110:9066-9071. doi: 10.1073/pnas.1219451110. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Karlsson CL, Onnerfält J, Xu J, Molin G, Ahrné S, Thorngren-Jerneck K. The microbiota of the gut in preschool children with normal and excessive body weight. Obesity (Silver Spring). 2012;20:2257-2261. doi: 10.1038/oby.2012.110. [DOI] [PubMed] [Google Scholar]
  • 18.Lim MY, You HJ, Yoon HS, et al. The effect of heritability and host genetics on the gut microbiota and metabolic syndrome. Gut. 2017;66:1031-1038. doi: 10.1136/gutjnl-2015-311326. [DOI] [PubMed] [Google Scholar]
  • 19.Plovier H, Everard A, Druart C, et al. A purified membrane protein from Akkermansia muciniphila or the pasteurized bacterium improves metabolism in obese and diabetic mice. Nat Med. 2017;23:107-113. doi: 10.1038/nm.4236. [DOI] [PubMed] [Google Scholar]
  • 20.Shin NR, Lee JC, Lee HY, et al. An increase in the Akkermansia spp. population induced by metformin treatment improves glucose homeostasis in diet-induced obese mice. Gut. 2014;63:727-735. doi: 10.1136/gutjnl-2012-303839. [DOI] [PubMed] [Google Scholar]
  • 21.Depommier C, Everard A, Druart C, et al. Supplementation with Akkermansia muciniphila in overweight and obese human volunteers: a proof-of-concept exploratory study. Nat Med. 2019;25:1096-1103. doi: 10.1038/s41591-019-0495-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Zhang T, Li Q, Cheng L, Buch H, Zhang F. Akkermansia muciniphila is a promising probiotic. Microb Biotechnol. 2019;12:1109-1125. doi: 10.1111/1751-7915.13410. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Chevalier C, Stojanović O, Colin DJ, et al. Gut microbiota orchestrates energy homeostasis during cold. Cell. 2015;163:1360-1374. doi: 10.1016/j.cell.2015.11.004. [DOI] [PubMed] [Google Scholar]
  • 24.Wasimuddin Schlaeppi K, Ronchi F, et al. Evaluation of primer pairs for microbiome profiling from soils to humans within the One Health framework. Mol Ecol Resour. 2020;20:1558-1571. doi: 10.1111/1755-0998.13215. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Callahan BJ, McMurdie PJ, Rosen MJ, Han AW, Johnson AJ, Holmes SP. DADA2: high-resolution sample inference from Illumina amplicon data. Nat Methods. 2016;13:581-583. doi: 10.1038/nmeth.3869. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Camacho C, Coulouris G, Avagyan V, et al. BLAST+: architecture and applications. BMC Bioinformatics. 2009;10:421. doi: 10.1186/1471-2105-10-421. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Lin H, Peddada SD. Analysis of compositions of microbiomes with bias correction. Nat Commun. 2020;11:3514. doi: 10.1038/s41467-020-17041-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Friedman J, Alm EJ. Inferring correlation networks from genomic survey data. PLoS Comput Biol. 2012;8:e1002687. doi: 10.1371/journal.pcbi.1002687. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Ognyanova K. Network analysis with R and igraph: NetSci X Tutorial. 2016. www.kateto.net/networks-r-igraph
  • 30.Kodera SM, Das P, Gilbert JA, Lutz HL. Conceptual strategies for characterizing interactions in microbial communities. Iscience. 2022;25:103775. doi: 10.1016/j.isci.2022.103775. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Foster KR, Schluter J, Coyte KZ, Rakoff-Nahoum S. The evolution of the host microbiome as an ecosystem on a leash. Nature. 2017;548:43-51. doi: 10.1038/nature23292. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Hall CV, Lord A, Betzel R, et al. Co-existence of network architectures supporting the human gut microbiome. iScience. 2019;22:380-391. doi: 10.1016/j.isci.2019.11.032. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Martin M, Dragoš A, Hölscher T, et al. De novo evolved interference competition promotes the spread of biofilm defectors. Nat Commun. 2017;8:15127. doi: 10.1038/ncomms15127. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Jones JM, Grinberg I, Eldar A, Grossman AD. A mobile genetic element increases bacterial host fitness by manipulating development. Elife. 2021;10:e65924. doi: 10.7554/eLife.65924. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Khazaei T, Williams RL, Bogatyrev SR, Doyle JC, Henry CS, Ismagilov RF. Metabolic multistability and hysteresis in a model aerobe-anaerobe microbiome community. Sci Adv. 2020;666:eaba0353. doi: 10.1126/sciadv.aba0353. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Wells CL, Maddaus MA, Reynolds CM, Jechorek RP, Simmons RL. Role of anaerobic flora in the translocation of aerobic and facultatively anaerobic intestinal bacteria. Infect Immun. 1987;55:2689-2694. doi: 10.1128/iai.55.11.2689-2694.1987. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Ferrer M, Martins dos Santos VA, Ott SJ, Moya A. Gut microbiota disturbance during antibiotic therapy: a multi-omic approach. Gut Microbes. 2014;5:64-70. doi: 10.4161/gmic.27128. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Vlckova K, Gomez A, Petrželková KJ, et al. Effect of antibiotic treatment on the gastrointestinal microbiome of free-ranging western lowland gorillas (Gorilla g. gorilla). Microb Ecol. 2016;72:943-954. doi: 10.1007/s00248-016-0745-5. [DOI] [PubMed] [Google Scholar]
  • 39.Shkoporov AN, Chaplin AV, Shcherbakova VA, et al. Ruthenibacterium lactatiformans gen. nov., sp. nov., an anaerobic, lactate-producing member of the family Ruminococcaceae isolated from human faeces. Int J Syst Evol Microbiol. 2016;66:3041-3049. doi: 10.1099/ijsem.0.001143. [DOI] [PubMed] [Google Scholar]
  • 40.Shin Y, Park S-J, Paek J, et al. Bacteroides koreensis sp. nov. and Bacteroides kribbi sp. nov., two new members of the genus Bacteroides. Int J Syst Evol Microbiol. 2017;67:4352-4357. doi: 10.1099/ijsem.0.002226. [DOI] [PubMed] [Google Scholar]
  • 41.Fang D, Shi D, Lv L, et al. Bifidobacterium pseudocatenulatum LI09 and Bifidobacterium catenulatum LI10 attenuate D-galactosamine-induced liver injury by modifying the gut microbiota. Sci Rep. 2017;7:8770. doi: 10.1038/s41598-017-09395-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.Chen Y, Yang B, Stanton C, et al. Bifidobacterium pseudocatenulatum ameliorates DSS-induced colitis by maintaining intestinal mechanical barrier, blocking proinflammatory cytokines, inhibiting TLR4/NF-kappaB signaling, and altering gut microbiota. J Agric Food Chem. 2021;69:1496-1512. doi: 10.1021/acs.jafc.0c06329. [DOI] [PubMed] [Google Scholar]
  • 43.De Filippis F, Paparo L, Nocerino R, et al. Specific gut microbiome signatures and the associated pro-inflammatory functions are linked to pediatric allergy and acquisition of immune tolerance. Nat Commun 2021;12:5958. doi: 10.1038/s41467-021-26266-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44.Magri G, Comerma L, Pybus M, et al. Human secretory IgM emerges from plasma cells clonally related to gut memory B cells and targets highly diverse commensals. Immunity. 2017;47:118-134.e8. doi: 10.1016/j.immuni.2017.06.013. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 45.Ianiro G, Tilg H, Gasbarrini A. Antibiotics as deep modulators of gut microbiota: between good and evil. Gut. 2016;65:1906-1915. doi: 10.1136/gutjnl-2016-312297. [DOI] [PubMed] [Google Scholar]
  • 46.Rashidi A, Ebadi M, Rehman TU, et al. Gut microbiota response to antibiotics is personalized and depends on baseline microbiota. Microbiome. 2021;9:211. doi: 10.1186/s40168-021-01170-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 47.Tilg H, Zmora N, Adolph TE, Elinav E. The intestinal microbiota fuelling metabolic inflammation. Nat Rev Immunol. 2020;20:40-54. doi: 10.1038/s41577-019-0198-4. [DOI] [PubMed] [Google Scholar]
  • 48.Suez J, Zmora N, Zilberman-Schapira G, et al. Post-antibiotic gut mucosal microbiome reconstitution is impaired by probiotics and improved by autologous FMT. Cell. 2018;174:1406-1423.e16. doi: 10.1016/j.cell.2018.08.047. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Data Availability Statement

The 16S rRNA sequencing data have been deposited to National Center for Biotechnology Information (NCBI) under the following BioProject ID: PRJNA929331. This study does not report the original code. Any additional information required to reanalyze the data reported in this article is available from the lead contact on request.

Articles from Bioinformatics and Biology Insights are provided here courtesy of SAGE Publications

RESOURCES