Document Type : Review Article

Author

Department of BioTechnology Vignan's Foundation for Science, Technology & Research Guntur, A.P, India

10.22034/jpbb.2022.346237.1024

Abstract

Background: The investigation of microbial community composition has been significantly influenced by technological innovations in the next-generation sequencing-based techniques. These techniques depend on 16S rRNA; in particular, have been frequently used to analyse the whole collection of bacteria present in a specific environment. As a response, a plethora of specific bioinformatic pipelines has been constructed for handling this data.
Methods and Analysis: Meta Genome Rapid Annotation Using Subsystem Technology (MG-RAST) and Quantitative Insights into Microbial Ecology (QIIME) are two publicly accessible metagenomic analysis tools utilized in several investigations. Here, we present a comparative study of the human vs. chimpanzee dataset using QIIME and MG-RAST to assess their accuracy in taxonomy and diversity analysis.
Discussion: Here, the taxonomic assignment was more accurate using QIIME, which assigned a considerably greater number of reads at the family level. As a consequence, QIIME produced a more accurate sequence file, which enhanced the diversity analysis outcome. Finally, while installing QIIME requires informatics expertise, it provides a wide range of metrics that are relevant for downstream applications and, more importantly, it is not dependent on server delays.

Graphical Abstract

Comparative gut Microbiome Evaluation of two Distinct Organisms: A Critical Review

Keywords

Main Subjects

  1. Introduction

The gut is home to the vast majority of human-associated microorganisms, where they perform a variety of activities in human metabolism. Given that the diverse genetic variety provided the microbes helps to broadcast metabolic activity by decomposing undigested food, producing essential vitamins and minerals, and promoting immune system physiological growth and maturation. As a result, many variables of human health, behaviour, and food have an impact on the makeup of gut microbial populations. Because of the interplay between hosts and microorganisms, it is frequently difficult to distinguish between cause and effect, that is, whether a particular phenotype resulted, it was caused by, a change in the microbiome, or whether differences are found from sources of variation. Alterations in microbiome makeup have been related to various diseases [1]. A decrease in microbial diversity in human microbiomes, for example, hypothesized to contribute to the rise in autoimmune and inflammatory illnesses found in industrialized civilizations. However, dysbiosis frequently arises secondary to the overall health state and is not necessarily a good sign of a specific illness. Direct and controlled trials are one approach to evaluate the causative involvement of the microbiota in disease pathology, but this is difficult, if not impossible, for most host organisms due to a combination of ethical norms and the difficulties of manipulating microbiomes. A comparative method, which evaluates the consistency of trends across populations or species that have developed separately, has been used as an alternative strategy for finding correlations between a particular host feature and certain components of the microbiome [2].

Over the last decade, we have learned a lot about how genetics, lifestyle, age, and some medical interventions all impact on composition of the human gut microbiome. Numerous research and subsequent reviews have focused on these characteristics. Many of the original connections between gut microbiome composition and a specific feature have been confirmed, while others have been refuted, based on comparative assessments of numerous geographically or genetically diverse cohorts. For example, even in the absence of direct experimental evidence, the correlation between a lactase non-persistence polymorphism and a higher abundance of Bifidobacterium observed in several human population cohorts that regularly consume milk, including UK twin, Hutterite, German, and Dutch cohorts indicate a causative effect. Furthermore, these Bifidobacterium in the gut may ferment lactose, which explains some non-persisters can metabolize lactose more efficiently than persisters. Extending a comparative approach to include the gut microbiomes of the other great apes aids in understanding what distinguishes humans and the mechanisms responsible for species similarities and differences. In this review, we present a mostly anthropocentric viewpoint and discuss how the gut microbiomes of other great ape species might disclose new information about humans [3]. We may discover more about how different variables impact microbiome composition and how the human microbiome has varied from that of our closest relatives by contrasting discoveries on humans with studies on closely related hosts. Current research on the implications of social interactions on microbiome similarity in chimp troops, for example, enriches prior results on human family groupings, and tracking microbial strains across great-ape species can lead to the distinct bacterial lineages that are suited to the human microbiome.

Incredible improvements in microbial ecology and our understanding of the human microbiome have been facilitated by high-throughput DNA sequencing methods combined with improved bioinformatics techniques. QIIME (Quantitative Insights into Microbial Ecology) is an open-source bioinformatics software package for microbial community analysis based on DNA sequence data that provides a single analysis framework for raw sequence data analysis via publication-quality statistical analyses and interactive visualizations. In this article, we illustrate the use of the QIIME pipeline to evaluate microbial communities collected from various places on the bodies of transgenic and wild-type mice, as measured by 16S rRNA gene sequences produced on the Illumina MiSeq platform. We offer our proposed pipeline for microbial community analysis and provide suggestions for making crucial decisions during the process. We provide examples of some of the studies that QIIME can do and explore how additional tools, such as MG-RAST, may be used to build on these analyses [4].

Advances in DNA sequencing technologies and the availability of culture-independent sequencing methods and software for analysing the massive amounts of data generated by these technologies have greatly improved our ability to characterize microbial communities in a wide range of environments. The human microbiota is a collection of bacteria that live in the human body. In our bodies, microbial cells outnumber human cells by a factor of up to ten to one. These microbial communities are important for human physiology and development, and dysbiosis is now linked to diseases like obesity and Crohn's disease. Evidence from transplants into germ-free mice implies that some of these relationships may be causal because some phenotypes can be conveyed by transferring the microbiota even including the transmission of human phenotypes into mice.

Microbes play a significant role in almost all ecosystems [5], the human body settings such as the skin or the gut. Because of the connection with human body habitats, numerous studies on microbial community composition have been conducted to analyse its role in various metabolic pathways and establish whether it is engaged in causing to avoiding particular clinical disorders. Such research might serve to elucidate the pathophysiology of certain diseases, as well as lead to the creation of novel disease indicators and/or treatment options. Several human illnesses are strongly associated with dysbiosis of particular microbial populations [6]. Because of technical advances in sequencing procedures, nearly all of the microorganisms from a particular environment can be analyzed in a single run, avoiding cultivation steps. In particular, procedures based on 16S rRNA next-generation sequencing, which enable high throughput microbial identification within a given metagenome, constitute a strong tool for studying the composition and richness of microbial communities [7]. The massive volume of next-generation metagenomic data created by such processes needs the use of bioinformatics tools capable of its analysis. Proper taxonomic classification of each microbe in a target environment is essential to assess the structure, biodiversity, richness, and role of the community resident in that environment [8, 12]. MetaGenome Rapid Annotation Using Subsystem Technology (MG-RAST) is a free (HTTP://metagenomics.nmpdr.org) completely automated system capable of processing metagenome sequencing data through sequence alignment, functional, and phylogenetic assignments, and comparative metagenomics [9]. Quantitative Insights Into Microbial Ecology (QIIME) is an open-source software pipeline (http://qiime.sourceforge.net/) that can perform a wide range of analyses on microbial communities starting from raw sequence data, such as sequence alignment, identification of operational taxonomic units (OTUs), elaboration of phylogenetic and taxon-based analysis of diversity within and between samples. Both tools have been used effectively to examine a large number of metagenomic 16S ribosomal RNA datasets by examining their capabilities in data management [10]. We conducted a comparative bioinformatic study of the same dataset using QIIME and MG-RAST to assess their taxonomic assignment accuracy. The effectiveness of these two well-established approaches in assigning sequence reads to microorganisms at various phylogenetic levels and assessing the variety and richness of microbial communities is reported here.

  1. Findings
  2. To Review the basic characteristic of a human vs. chimpanzee gut microbiome study, determine how much DNA similarity exists.
  3. To investigate the common ancestor and characteristics of humans and chimpanzees of gut microbiomes using phylogenetic, diversity, taxonomic, and statistical studies.
  4. Methods and Analysis

3.1. MG-RAST

MG-RAST is a free and open-source web application server that suggests the automated phylogenetic and functional analysis of metagenomes. It is also one of the largest metagenomic data sources [11]. The program generates functional designations for metagenome sequences automatically by comparing them to databases at both the nucleotide and amino acid levels. The program provides phylogenetic and functional designations for the metagenome under study, as well as tools for comparing metagenomes.

3.2. Background of MG-RAST

MG-RAST was created to free, a public resource for analysing and storing metagenome sequencing data. The service eliminates one of the most significant barriers in metagenome analysis: the availability of high-performance computers for data annotation. Because metagenomic and metatranscriptomic research entails the processing of enormous datasets, they often necessitate computationally intensive analysis. Scientists may now create such large amounts of data because sequencing prices have dropped drastically in recent years. This has moved the limiting issue to computer costs: for example, a recent University of Maryland research predicted a cost of more than $5 million per terabase utilizing their CLOVR metagenome analysis pipeline. As the size and number of sequencing databases grow, so will the expenses associated with their analysis [12]. MG-RAST likewise works as a vault instrument for metagenomic information. Metadata assortment and understanding are basic for genomic and metagenomic exploration, and issues in this space incorporate data trade, curation, and circulation. The MG-RAST framework was an early adopter of the Genomics Requirements Consortium's essential agenda principles and improved biome-explicit ecological bundles, and it incorporates a simple to-utilize uploader for metadata gathering at the hour of information accommodation [13, 24].

3.3. Analysis framework for metagenomic samples

Using a variety of bioinformatics tools, the MG-RAST program provides automatic quality checking, annotation, comparative analysis, and archiving of metagenomic and amplicon sequences. The program was designed to examine metagenomic data, but it also supports the analysis of amplicon (16S, 18S, and ITS) sequences and metatranscriptome (RNA-seq) sequences [14, 25]. MG-RAST is currently incapable of predicting coding regions from eukaryotes, making it ineffective for analysing eukaryotic metagenomes. The MG-RAST pipeline is broken into five stages:

  1. Data scrubbing

Steps for quality control and artifact removal are included. First, low-quality areas are edited with SolexaQA and readings with excessive lengths are deleted. When processing metagenome and metatranscriptome datasets, a de-replication phase is implemented. Following that, DRISEE (Duplicate Read Inferred Sequencing Error Estimation) is utilized to calculate the sample sequencing error using Artificial Duplicate Reads (ADRs). Finally, the pipeline allows for the screening of reads using the Bowtie aligner and the removal of reads that exhibit near matches to model organism genomes (including fly, mouse, cow, and human).

  1. Extraction of features

C.FragGeneScan, a machine learning method, is used by MG-RAST to identify gene sequences. An initial BLAST search against a truncated version of the SILVA database yields ribosomal RNA sequences.

  1. Annotation of features

MG-RAST uses the UCLUST implementation in QIIME to generate clusters of proteins with 90% similarity to discover the likely functions and annotation of the genes. A similarity analysis will be performed on the longest sequence in each cluster. The similarity analysis is performed using sBLAT (the BLAT technique is parallelized using OpenMP). The search is performed against a protein database constructed from the M5nr, which incorporates non-redundant sequences from the GenBank, SEED, IMG, UniProt, KEGG, and eggNOGs databases. At 97 percent identity, the reads associated with rRNA sequences are grouped. The longest sequence from each cluster is chosen as the representative and utilized in a BLAT search against the M5rna database, which incorporates SILVA, Greengenes, and RDP.

  1. Profile creation

Profiled to generate the data is used in a variety of data packages. The abundance profiles, which are rotated and aggregated versions of the similarity files, are the most essential.

  1. Data loading

Finally, the abundance profiles are put into the appropriate databases.

3.4. QIIME

QIIME is a bioinformatic pipeline intended to break down microbial networks assembled utilizing marker quality amplicon sequencing. At its centre, the pipeline performs quality control on the info sequencing peruses, bunches the marker quality nucleotide groupings at an ideal phylogenetic level into OTUs or succession variations, and systematically clarifies by the scanning a reference ordered data set for comparative sequences[15, 26]. The QIIME work process produces an element table that sums up the wealth of each OTU or grouping variety in each example. Inside the pipeline, different instruments connected with the natural highlights of the examples being explored are likewise offered, for example, rarefaction, alpha variety and beta variety estimations, perceptions, for example, chief directions examination (PCoA), and considerably more. QIIME adopts a somewhat secluded strategy and takes into consideration the utilization of numerous procedures at different phases of the study [29]. For instance, the step of succession grouping can be directed with UCLUST, CD-HIT, BLAST, and different devices. QIIME has been effectively evolved since its underlying delivery in 2010 [30].

3.5. QIIME as A Third-Party Service Involved in Workflow

The trouble of introducing QIIME, due to a limited extent to the tremendous number of programming conditions, was an early obstacle to acknowledgment. The large number of conditions, then again, was a purposeful choice taken during QIIME development [16]. To build a pipeline for succession investigation that incorporates many advances like grouping assortment, curation, and factual examination, the client should consider many existing instruments that have been created to carry out unambiguous roles and widely benchmarked on their capacity to carry out these roles, for example, the bunch program for bunching arrangements into Operational Taxonomic Units. As a premise, a pipeline has two choices: re-implement the calculation or influence existing programming. As opposed to re-implementing the calculations, the QIIME engineers decided to altogether embody them. This choice keeps up with the respectability of the pipeline's projects since no question that the apparatus is being utilized according to plan, fabricated, and tried by the first creators and, as a rule, peer-investigated by the logical community [17, 35]. Despite the fact that current programming is reused, the QIIME pipeline can incorporate and scatter newfound and further developed calculations more rapidly than if every calculation must be re-implemented and retested to guarantee that it matched the first. As a thought, QIIME clients might be certain that they are involving the most exceptional instruments for their investigations and can appropriately recognize the makers of the part programming bundles [18].

3.6. QIIME Workflow for Undertaking Microbial Community Evaluation

In a single run, the Illumina MiSeq technology may create up to 107 sequences. QIIME analyses instrument data to produce meaningful information about the community represented in each sample. This process is coarsely divided into "upstream" and "downstream" stages. The upstream stage covers all raw data processing and the creation of crucial files for the microbiological investigation. The downstream stage performs diversity analysis, statistics, and interactive visualizations of the data using the OTU table and phylogenetic tree created in the upstream step. Furthermore, QIIME is rapidly interfacing with other programs like as IPython and R, allowing for further analysis [19].

3.7. Procedures in the prior analysis: An QIIME assessment

The procedure begins with the sequence and an aligning file created by the user. The aligning file, which is in tab-delimited text format, includes information for comprehending what is in each sample and is thus crucial for executing the remaining studies. The primary records on this document is a unique identifier for every pattern, the barcode used for every pattern, the primer series used, and a description for each sample, as well as extra consumer-defined information required for publishing the implications, together with which species the sample become taken from which site on the body is being studied, and clinical variables relevant to the study [20].

The initial phase of the QIIME method requires the identification, characterization, labeling, and primer sequence. This stage comprises sample identification, characterization, and quality screening. Additional records about the samples supplied in the aligning file is useful for subsequent phases, particularly studies that aggregate the samples based on these fields. As a result, we advocate adding as much extra data on the samples as possible. This additional information can be further used to identify tainted samples. SourceTracker, for example, is a Depending on a library of data from recognized communities; QIIME software determines the proportion of varied community sources, especially infection, in each sample.

Although the digestive tract microbial neighbourhoods of the chimpanzees analysed are actually compositionally more just like each other compared to they are actually to the digestive tract microbial neighbourhoods of people (p = 1.029 × 10−86, one-tailed t-test; Additional Number S1), all of 8 of the microbial genera that are actually distinctively overrepresented in each an individual and a chimpanzee enterotype reveal the exact very same wealth designs throughout enterotypes in each multitude types. Each individual enterotype 1 as well as chimpanzee enterotype 1 are actually overrepresented through Bacteroides, Faecalibacterium, and Parabacteroides.

  1. Discussion

The characteristics of our closest evolutionary cousins were reviewed. Chimpanzees are our closest living cousins, with whom we shared a common ancestor just a million years ago [21, 41]. The DNA identity and total genome content similarity between humans and chimps approach 98.1%. To some extent, though, this close kinship is mirrored in their gut microbiomes. The collection of chimp microbiomes is advantageous in that most samples were collected before microbiome analysis became standard, before the current broad attempts to identify the function of the microbiome in physiology and health [22,45]. Chimpanzees are well represented by samples from several individuals of known age, diet, provenance, location, gender, ancestry, seasonal eating habits, migration patterns, social interactions, and health state, to continue anthropological, behavioral, and epidemiological investigations. Although current human microbiome sampling significantly exceeds that of chimps, these preserved records permitted many of the early analyses of genetic, ecological, and social variables that affect the makeup of gut microbiomes [23, 50].

This review aimed to assess the efficacy of bioinformatic analyses of 16S rRNA next-generation sequencing-based data conducted by QIIME and MG-RAST, two of the most commonly mentioned tools in the context of a metagenomic investigation. We began by assessing the accessibility and usability of both technologies. QIIME is an open-source software package, whereas MG-RAST is a web platform for automated analysis. In the first scenario, the analysis is dependent on the MG-RAST server's timings and data uploading constraints. The user should submit the raw data to the MG-RAST server, stating whether the data is secret (accessible only to the submitter) or public (available to everyone) (data will be shared with all MG-RAST users).MG-RAST offers five choices with varying analysis timeframes, each connected with a priority queue. We select "The Lowest Priority" (data will stay secret) for scientific research objectives. The time required by the MG-RAST server to finish the analysis is proportional to the number of jobs put in the analysis queue by all MG-RAST users and the priority level chosen. This may not always be consistent with the researcher’s demands [24].

However, various dependencies should be installed individually to access the entire QIIME analysis pipeline and installation needs some basic informatics expertise. To compensate for this limitation, QIIME provides some free choices [53-55]. Users can begin the analytic process after installing QIIME and all of its dependencies. The time required to finish installation is determined by numerous factors, including the amount of data, the pipeline used, and the user's bioinformatics expertise. We chose the same settings for the preliminary analysis to minimize disparities between MG-RAST and QIIME. Quality filtering, primer identification, and read de-multiplexing are all part of this stage. MG-RAST does analysis, whereas QIIME combines the data into a script (split libraries.py). Both programs need a metadata aligning file for the analysis filtering stage, this requires the user to provide at least adequate data: (i) sample Identifier and barcode, (ii) primer patterns utilised in library construction, and (iii) one or more description fields including test statistics. Both techniques detected six distinct Phyla with comparable identities in our sample. Surprisingly, discrepancies in taxonomic identification at the family level were statistically significant [25-30]. QIIME, in particular, allocated readings more precisely to distinct families, while a lesser percent of reads were assigned to the kinds "No Hits" and "Unclassified." Following taxonomic venture, which gives a picture of microbial community composition, a metagenomic analysis pipeline generally maintains to assess microbial variety as alpha diversity (quantitative global diversity inside a sample) and beta diversity (qualitative variety between a set of samples)[31-34].

Metrics are important for estimating communal wealth and determining the microbial compositional resemblance between samples. In terms of alpha diversity calculated using multiple metrics, we achieved similar findings with MG-RAST and QIIME. However, while doing beta diversity analysis, various values were assigned to the same topic depending on the tool used, even though they were generated using the same measure. As a review, this disparity may be due to discrepancies in 16S rRNA identification and taxonomy classification [35-40].

  1. Conclusion

We performed a comparative metagenomic investigation of the gut microbiome composition in the Human Vs Chimpanzee using both the MG-RAST and QIIME pipelines. Our results show that the QIIME technique provides more exact taxonomic identification, which is important for future diversity assessments. Furthermore, because it is freely available for download, it is not reliant on server timings. Finally, QIIME incorporates the BIOM file directly into its pipeline, which is useful for a number of downstream analyses, while also expediting the entire process. Less experienced operators, on the other hand, may find MG-RAST easier to use than QIIME. Given all of the aforementioned features, we feel that MG-RAST might be useful for initial users learning about metagenomic analysis findings and criticisms. In the following year, upgraded versions of QIIME will be launched with even more capabilities, including a User-Interface to help non-computer-skilled individuals easily analyse their data. Figure 1 depicts the proposed diagram, a useful tool for microbiome analysis. It allows scientists to retrace the critical phases in sequence processing from raw data to visualisations and interpretation of the findings.

QIIME is particularly beneficial because of two advantages: faithfulness to the methods utilized and accuracy in the evaluation. Because QIIME covers present software program, the integrity of the authentic programs and algorithms planned, created, and tested by way of the unique developers is preserved. Due to the fact QIIME may be applied to sequences from many platforms, after the upstream manner is finished, the evaluation (downstream) step is the same irrespective of the sequencing platform utilized. Those features, together with the truth that QIIME is open supply software program with consistent aid for customers thru the QIIME discussion board, have contributed to the fast increase of the QIIME user network concerning its launch. QIIME implements downstream and upstream strategies in a way that offers numerous options for analysing.

We examine and illustrate the ideas for each stage, what the scripts do, and how to pick amongst alternatives in this review. This article offers an overview of several of the common procedures in a microbiota study based on the examination of 16S rRNA sequences obtained by high-throughput sequence analysis, irrespective of the usage of QIIME. Some of these tools have a long history in general ecology, while others are still in the early stages of development; we encourage microbial ecologists and data scientists to collaborate to create, develop, and implement new techniques and tactics that will allow even more investigations of this captivating field.

  1. Acknowledgements

The authors appreciate Agri, its solutions' technical assistance, and the fruitful talk of insightful suggestions concerning QIIME. They also thank Dr. Sukesh Kalva and Dr. Tarun Pal for assisting in the collection of the dataset and enabling us to utilize it. This activity was partly supported by Vignan University.

Authors’ Contributions

Venu Parital poses the analysis was conceived of and created by myself. Data collection; data or analytic tools were contributed, carried out the analysis and wrote the article .

Conflict of interest

The authors state that they have no conflicts of interest.

Financial support

None.

Ethics statement

None.

Orcid

Venu Paritala: https://orcid.org/0000-0003-0385-5322

  1. Baker B J, Sheik C S, Taylor C A, et al., (2013). Community transcriptomic assembly reveals microbes that contribute to deepsea carbon and nitrogen cycling, The ISME Journal, 7: 1962–1973. ‎[Crossref], [Google Scholar], [Publisher]‎
  2. Philippot L, Raaijmakers J M, Lemanceau P, van der Putten W H. (2013‎). Going back to the roots: the microbial ecology of the rhizosphere, Nature Reviews Microbiology, 11: 789–799. ‎[Crossref], [Google Scholar], [Publisher]‎
  3. Zeeuwen P L, Kleerebezem M, Timmerman H M, Schalkwijk J. (2013). Microbiome and skin diseases, Current Opinion in Allergy and Clinical Immunology, 13: 514–520. ‎[Crossref], [Google Scholar], [Publisher]‎
  4. Cho I, Blaser M J. (2012). The human microbiome: at the interface of health and disease, Nature Reviews Genetics, 13: 260–270. ‎[Crossref], [Google Scholar], [Publisher]‎
  5. Fang S, Evans R M. (2013). Microbiology: wealth management in the gut, Nature, 500: 538–539. ‎[Crossref], [Google Scholar], [Publisher]‎
  6. Knights D, Lassen K G, Xavier R J. (2013). Advances in inflammatory bowel disease pathogenesis: linking host genetics and the microbiome, Gut, 62: 1505–1510. ‎[Crossref], [Google Scholar], [Publisher]‎
  7. Wu N., Yang X., Zhang R. et al. (2013). Dysbiosis signature of fecal microbiota in colorectal cancer patients, Microbial Ecology, 66: 462–470. ‎[Crossref], [Google Scholar], [Publisher]‎
  8. D’Argenio V, Precone V, Casaburi G. et al. (2013). An altered gut microbiome profile in a child affected by Crohn’s disease normalized after nutritional therapy, American Journal of Gastroenterology, 108: 851–852. ‎[Crossref], [Google Scholar], [Publisher]‎
  9. Gollwitzer E S, Marsland B J. (2014). Microbiota abnormalities in inflammatory airway diseases—potential for therapy, Pharmacology & Therapeutics, 141: 32–39. ‎[Crossref], [Google Scholar], [Publisher]‎
  10. Kim B S, Jeon Y S, Chun J. (‎2013‎). Current status and future promise of the human microbiome, Pediatric Gastroenterology, Hepatology & Nutrition, 16: 71–79. ‎[Crossref], [Google Scholar], [Publisher]‎
  11. Ribeca P, Valiente G. (2011). Computational challenges of sequence classification in microbiomic data, Briefings in Bioinformatics, 12(6): 614–625. ‎[Crossref], [Google Scholar], [Publisher]‎
  12. De Filippo C, Ramazzotti M, Fontana P, Cavalieri D. (2012). Bioinformatic approaches for functional annotation and pathway inference in metagenomics data, Briefings in Bioinformatics, 13: 696–710. ‎[Crossref], [Google Scholar], [Publisher]‎
  13. Meyer F, Paarmann D, D’Souza M. et al., (‎2008‎). The metagenomics RAST server—a public resource for the automatic phylogenetic and functional analysis of metagenomes, BMC Bioinformatics, 9: 386. ‎[Crossref], [Google Scholar], [Publisher]‎
  14. Caporaso J G, Kuczynski J, Stombaugh J. et al. (2010). QIIME allows analysis of high-throughput community sequencing data, Nature Methods, 7(5): 335–336. ‎[Crossref], [Google Scholar], [Publisher]‎
  15. Teeling H, Glockner F O. (2012). Current opportunities and ¨ challenges in microbial metagenome analysis: a bioinformatic perspective, Briefings in Bioinformatics, 13: 728–742. ‎[Crossref], [Google Scholar], [Publisher]‎
  16. Navas-Molina J A, Peralta-Sanchez J M, Gonz ´ alez A. et al., (2013). Advancing our understanding of the human microbiome using QIIME, Methods in Enzymology., 531: 371–444. ‎[Crossref], [Google Scholar], [Publisher]‎
  17. Kent W J. (2002). BLAT—the BLAST-like alignment tool, Genome Research, 12(4): 656–664. ‎[Crossref], [Google Scholar], [Publisher]‎
  18. Pruesse E, Quast C, Knittel K. et al. (2007). SILVA: a comprehensive online resource for quality checked and aligned ribosomal RNA sequence data compatible with ARB, Nucleic Acids Research, 35(21): 7188–7196. ‎[Crossref], [Google Scholar], [Publisher]‎
  19. Edgar R C, (2010). Search and clustering orders of magnitude faster than BLAST,” Bioinformatics, 26(19): 2460–2461. ‎[Crossref], [Google Scholar], [Publisher]‎
  20. McDonald D, Price M N, Goodrich J. et al. (2012). An improved Greengenes taxonomy with explicit ranks for ecological and evolutionary analyses of bacteria and archaea, The ISME Journal, 6(3): 610–618. ‎[Crossref], [Google Scholar], [Publisher]‎
  21. McDonald D, Clemente J C, Kuczynski J. et al. (2012). The Biological Observation Matrix (BIOM) format or: how I learned to stop worrying and love the ome-ome, GigaScience, 1(1). ‎[Crossref], [Google Scholar], [Publisher]‎
  22. Altschul S F, Gish W, Miller W, Myers E W, Lipman D J. (1990). Basic local alignment search tool, Journal of Molecular Biology, 215(3): 403–410. ‎[Crossref], [Google Scholar], [Publisher]‎
  23. Wang Q, Garrity G M, Tiedje J M, Cole J R, (2007). Na¨ıve Bayesian classifier for rapid assignment of rRNA sequences into the new bacterial taxonomy, Applied and Environmental Microbiology, 73(16): 5261–5267,. ‎[Crossref], [Google Scholar], [Publisher]‎
  24. Hill T C J, Walsh K A, Harris J A, Moffett B F. (2003). Using ecological diversity measures with bacterial communities, FEMS Microbiology Ecology, 43(1): 1–11. ‎[Crossref], [Google Scholar], [Publisher]‎
  25. Spellerberg I F, Fedor P J. (2003). A tribute to Claude-Shannon (1916-2001) and a plea for more rigorous use of species richness, species diversity and the “Shannon-Wiener” Index,” Global Ecology & Biogeography, 12(3): 177–179. ‎[Crossref], [Google Scholar], [Publisher]‎
  26. Beals E W. (1984). Bray-Curtis ordination: an effective strategy for analysis of multivariate ecological data, Advances in Ecological Research, 14: 1–55. ‎[Crossref], [Google Scholar], [Publisher]‎
  27. Field A. (2007). Analysis of variance (ANOVA), in Encyclopedia of Measurement and Statistics, N. Salkind, Ed., SAGE Publications, Thousand Oaks, Calif, USA. ‎[Crossref], [Google Scholar], [Publisher]‎
  28. Sokal R R, Rohlf F J. (1981). Biometry: The Principles and Practice of Statistics in Biological Research, Freeman, New York, NY, USA. ‎[Crossref], [Google Scholar], [Publisher]‎
  1. Abdi H. (‎2007‎). Bonferroni and Sid ˇ ak corrections for multiple com- ´ parisons, in Encyclopedia of Measurement and Statistics, N. J. Salkind, Ed., SAGE Publications, Thousand Oaks, Calif, USA. ‎[Crossref], [Google Scholar], [Publisher]‎
  2. Bagdonavicius V, Kruopis J, Nikulin M S. (‎2011‎). Non-Parametric Tests for Complete Data, John Wiley & Sons, London, UK. ‎[Crossref], [Google Scholar], [Publisher]‎
  3. Digby ‎P G N, Kempton R A. (1987) Multivariate Analysis of Ecological Communities, Chapman & Hall, London, UK. ‎[Crossref], [Google Scholar], [Publisher]‎
  4. Anderson M J. (2001). A new method for non-parametric multivariate analysis of variance,” Austral Ecology, 26: 32–46. ‎[Crossref], [Google Scholar], [Publisher]‎
  5. Crawley M J. (2002) Statistical Computing: An Introduction to Data Analysis Using S-PLUS, John Wiley & Sons. ‎[Crossref], [Google Scholar], [Publisher]‎
  6. Schloss P D, Westcott S L, Ryabin ‎ et al., (2009). Introducing mothur: open-source, platform-independent, communitysupported software for describing and comparing microbial communities, Applied and Environmental Microbiology, 75: 7537–7541. ‎[Crossref], [Google Scholar], [Publisher]‎
  7. Soergel D A W, Dey N, Knight R, Brenner S E. (2012). Selection of primers for optimal taxonomic classification of environmental 16S rRNA gene sequences, The ISME Journal, 6: 1440–1444. ‎[Crossref], [Google Scholar], [Publisher]‎
  8. Quast C, Pruesse E, Yilmaz P. et al. (2013). The SILVA ribosomal RNA gene database project: improved data processing and webbased tools, Nucleic Acids Research, 41: D590–D596. ‎[Crossref], [Google Scholar], [Publisher]‎
  9. Whittaker R H. (1972). Evolution and measurement of species diversity, Taxon, 21: 213–251. ‎[Crossref], [Google Scholar], [Publisher]‎
  10. Allaire J, Horner J, Marti V, Porte N. (2013). Markdown: Markdown rendering for R. ‎[Crossref]
  11. Baulcombe DC, Bäumlein H, Wobus U, Pustell J, Kafatos FC, Bendich A, Benfey PN, Takatsuji H, Ren L, Shah DM, Chua NH. Altschul, sf, gish, w., miller, w., myers, ew & lipman, dj (1990). J. Mol. Biol. 1998;215:403-10, [Google Scholar], [Publisher]‎
  12. Arumugam M, Raes J, Pelletier E, Le Paslier D, Yamada T, Mende D R. et al. (2011). ‎Basic local alignment search tool. Journal of Molecular Biology, 215(3): 403–410. ‎[Crossref], [Google Scholar], [Publisher]‎