A guide for comparing microbial co‐occurrence networks

iMeta, Journal Year: 2023, Volume and Issue: 2(1)

Published: Jan. 10, 2023

The article provides a pipeline for comparing microbial co-occurrence networks based on the R microeco package and meconetcomp package. It has high flexibility expansibility can help users efficiently compare built from different groups of samples or construction approaches. Microorganisms are ubiquitous in diverse environments earth play important roles ecosystem functions ranging biogeochemical cycles [1] to maintenance host health [2, 3]. Microbial assemblages generally comprised large number species, which is represented as "microbial community" defined context spatiotemporal scales. Identifying members their abundance community basic task ecology studies. Over past two decades, giant leap forward sequencing techniques made this possible, leading rapid increase data size. Furthermore, advances bioinformatic softwares (e.g., QIIME2 [4]) have profoundly improved speed convenience sequence analysis. After obtaining operational taxonomic units (OTU), amplicon variants (ASV) species abundances (sequence counts estimated abundance) analysis process, following statistics visualizations be performed using language [5] related packages, grown up cutting-edge system recent decades [6]. Many statistical approaches used microbiome benefiting similarities between macro- micro-ecosystems. However, there also some dissimilarities research methods routes these ecosystems [2]. invisibility microbes, proportion uncultured huge diversity lead difficulty hypothesis-driven studies, especially those referring interactions functions. Researchers usually need try series tools find suitable one verify hypothesis. To maximize accessibility provide good user experience, code must organized manner that adheres both conciseness functionality. Based background, [7] was developed R6 class make customized easier faster. In addition, file2meco [8] facilitate conversion files other software, such [4] HUMAnN [9]. network often decipher hidden patterns complex consortia wide range studies (see [10] references therein). contrast macro-organisms (mainly observations [11, 12]), mostly constructed count tables obtained metagenomic data. several general issues rendering challenge, including compositionality denote proportions instead absolute abundances), sparsity (a zeros), inference direct associations paired taxa. Another challenge how explain edges signs given not recommended interpreted cross-sectional [13]. best our knowledge, correlation-based (Pearson Spearman correlation) may earliest widely approach studied habitats, soil [14]. address existed correlation network, been developed, Sparse Correlations Compositional (SparCC) [15], Compositionally Corrected by REnormalization PErmutation (CCREPE) [16], Correlation through Lasso (CCLasso) [17]. Further, graphical model created robustly infer taxa optimize structure. For example, SParse InversE Covariance Estimation Ecological Association Inference (SPIEC-EASI) [18] combines transformations compositional algorithms sparse neighborhood inverse covariance selection reconstruct network. FlashWeave [19] adopts local-to-global learning framework directly associated neighbors (i.e., Markov blanket) taxon scalability heterogenous sets. comparisons [20] reviews thoroughly discussed robustness particularly depending upon challenges. BEEM-static method [21] dedicated seek out with generalized Lotka-Volterra (gLV) an expectation-maximization algorithm, offering directed gain insight into communities. Along development, controversial voices worry about misuse network-related (especially network) biotic [13, 22, 23]. main reason hold hub answering many questions ecology. that, no matter used, contain more less information approaches, parameters, features themselves. There another case actual captured because itself biological characteristics higher-order interaction [24, 25]). Broadly speaking, interpretations represent largely overlooked mainly due its difficulty. Although recently reveal among microbes [19, 21], applicability still clear when heterogeneous sets applied. So it 26], producing dilemma what extent interactions. appealed inferred background knowledge previous 26]. Even macroecology empirical information, study revealed rarely matched net [27], highlighting inadvisable practice coupling co-occurrences terms ecological hypothesis, currently frequently null constrained ordination conditional deterministic consortia, random (such drift dispersal neutral process) cannot generate strong patterns. understood new application classic "checkerboard distribution" approach. derived joint distribution historical legacy. Various layers complexity inherent systems variability multiple abiotic factors factors) blur whether association real influence association. Moreover, correspond dependent models underlying algorithms, generating problem interpretation. prevalence boosted burst various fields, far enough hypotheses mechanisms behind descriptive metrics. implemented software "black-box" lack combine types methods. reasonably use researchers thorough understanding flows details construction. practice, improve edge finding. complicated experimental design, treatments [28], urgency user-friendly comparison. already integrated online MetagenoNets [29] iNAP [30]) packages igraph [31], NetCoMi [32], ggClusterNet [33]) devoted visualizations. But comparison daunting challenge. protocol, comparison, (https://github.com/ChiLiubio/meconetcomp) (Figure 1A). protocol introduces usage trans_network much possible classes show power 1B). A characterized edge-weighted graph G = (V, E), where V (node) represents feature (ASV/OTU/species) E (edge) encodes connection weight denotes strength connection. (+ −) positive negative associations. set soil_amp stool_met microtable objects prepared. How import own users? Please read document tutorial (https://chiliubio.github.io/microeco_tutorial/basic-class.html#microtable-class). Pearson datasets, adding parameter "use_WGCNA_pearson_spearman TRUE" calculation. Note requires WGCNA installed (https://cran.r-project.org/web/packages/WGCNA/). parameters construction, please see function cal_network command: help(trans_network). Module hubs (nodes highly links within module, Zi > 2.5 Pi ≤ 0.62); (2) Connectors connect modules, (3) Network act module connectors, (4) Peripherals only few almost always nodes < 0.62). we introduced Different take full advantage list class, perform each part most noteworthy limitation practice. applied covered protocol. date, standards available Nor "one size fits all" tool automatically match user's set. Similar results benchmark detection strategies vary sensitivity precision [20], found across at aspect 2B saved figures stool_met_network set). We argued except selections [10, 20, 25], feasible valuable learn reflect integrating model) packages. Disentangling environmental effects attractive topic [39]. Yet gold standard diminish feasibility studying questions. report shows rare likely caused metabolic cross-feeding [40]. habitats like soil, should interpretation interacts. though determinate impede methodological development limit ability broadly link patterns, assembly processes combining instance, differential phylogenetic distance 2C) showed (strong correlations) might varying processes. affect topological structure soils categories Chinese wetlands 2E). With increasing cultured metagenome-assembled genomes, meaning shared 2B) figure interact [41] explaining results. While all beyond scope examples demonstrated generated combination list, "for" loop, easy use, did consider too manipulations, rarefaction, filtering, parameters. These operations critical sense they expected. since object special attribute binding objects, loop package, classes. All authors contributed development. initial idea conceived Chi Liu, Minjie Yao, Xiangzhen Li. During test improvements. original manuscript written Li, revised Chaonan Yanqiong Jiang, Raymond J. Zeng. This work supported National Natural Science Foundation China (42077206 32070548). thank four anonymous reviewers helpful suggestions manuscript. reporting bugs, suggestions, usability problems. declare conflict interest. Besides datasets loaded deposited GitHub (https://github.com/ChiLiubio/network_protocol) Gitee (https://gitee.com/chiliubio/network_protocol). Supporting Information materials (figures, tables, scripts, abstracts, slides, videos, translated versions, update materials) DOI iMeta http://www.imeta.science/.

Language: Английский

---

Depth effects on bacterial community altitudinal patterns and assembly processes in the warm-temperate montane forests of China

The Science of The Total Environment, Journal Year: 2024, Volume and Issue: 914, P. 169905 - 169905

Published: Jan. 6, 2024

Language: Английский

---

Microplastics enhance the invasion of exotic submerged macrophytes by mediating plant functional traits, sediment properties, and microbial communities

Journal of Hazardous Materials, Journal Year: 2024, Volume and Issue: 469, P. 134032 - 134032

Published: March 13, 2024

Language: Английский

---

Aerobic composting with sauerkraut fermentation waste water: Increasing the stability and complexity of bacterial community and changing bacterial community assembly processes

Bioresource Technology, Journal Year: 2023, Volume and Issue: 376, P. 128883 - 128883

Published: March 14, 2023

Language: Английский

---

Nutrient-dependent cross-kingdom interactions in the hyphosphere of an arbuscular mycorrhizal fungus

Frontiers in Microbiology, Journal Year: 2024, Volume and Issue: 14

Published: Jan. 4, 2024

Introduction The hyphosphere of arbuscular mycorrhizal (AM) fungi is teeming with microbial life. Yet, the influence nutrient availability or forms on microbiomes still poorly understood. Methods Here, we examined how community (prokaryotic, fungal, protistan) was affected by presence AM fungus Rhizophagus irregularis in rhizosphere and root-free zone, different nitrogen (N) phosphorus (P) supplements into compartment influenced communities. Results greatly communities both prokaryotic being most. Protists were only group microbes whose richness diversity significantly reduced fungus. Our results showed that type nutrients encounter localized patches modulate structure In contrast did not observe any effects (non-mycorrhizal) fungal composition. Compared to non-mycorrhizal control, zone (i.e., hyphosphere) enriched Alphaproteobacteria , some micropredatory copiotroph bacterial taxa (e.g., Xanthomonadaceae Bacteroidota ), characterized yet cultured Acidobacteriota subgroup GP17, especially when phytate added. Ammonia-oxidizing Nitrosomonas nitrite-oxidizing Nitrospira suppressed compartment, upon addition inorganic N. Co-occurrence network analyses revealed complex interconnected more keystone species present, amended phytate. Conclusion study form an important driver eukaryotic assembly hyphosphere, despite assumed a stable specific hyphoplane microbiome. Predictable responses will open possibility using them as co-inoculants fungi, e.g., improve crop performance.

Language: Английский

---

Role of indigenous microbial communities in the mobilization of potentially toxic elements and rare-earth elements from alkaline mine waste

Journal of Hazardous Materials, Journal Year: 2024, Volume and Issue: 466, P. 133504 - 133504

Published: Jan. 11, 2024

Language: Английский

---

Habitat-specific environmental factors regulate the spatial variability of biological soil crust microbial communities on the Qinghai-Tibet Plateau

The Science of The Total Environment, Journal Year: 2023, Volume and Issue: 901, P. 165937 - 165937

Published: July 31, 2023

Language: Английский

---

Ketogenic diet therapy for pediatric epilepsy is associated with alterations in the human gut microbiome that confer seizure resistance in mice

Cell Reports, Journal Year: 2023, Volume and Issue: 42(12), P. 113521 - 113521

Published: Dec. 1, 2023

The gut microbiome modulates seizure susceptibility and the anti-seizure effects of ketogenic diet (KD) in animal models, but whether these relationships translate to KD therapies for human epilepsy is unclear. We find that clinical alters microbial function children with refractory epilepsy. Colonizing mice KD-associated microbes promotes resistance relative matched pre-treatment controls. Select metagenomic metabolomic features, including those related anaplerosis, fatty acid β-oxidation, amino metabolism, are seen therapy preserved upon transfer mice. Mice colonized exhibit altered hippocampal transcriptomes, pathways ATP synthesis, glutathione oxidative phosphorylation, linked genes identified Our findings reveal key functions by pediatric microbiome-induced alterations brain gene expression protection

Language: Английский

---

Stepwise degradation of organic matters driven by microbial interactions in China΄s coastal wetlands: Evidence from carbon isotope analysis

Water Research, Journal Year: 2023, Volume and Issue: 250, P. 121062 - 121062

Published: Dec. 23, 2023

Language: Английский

---

Biological nitrogen fixation and the role of soil diazotroph niche breadth in representative terrestrial ecosystems

Soil Biology and Biochemistry, Journal Year: 2023, Volume and Issue: 189, P. 109261 - 109261

Published: Nov. 30, 2023

Language: Английский

---