Unravelling the enigma of the human microbiome: Evolution and selection of sequencing technologies

Abstract

The human microbiome plays a crucial role in maintaining health, with advances in high-throughput sequencing technology and reduced sequencing costs triggering a surge in microbiome research. Microbiome studies generally incorporate five key phases: design, sampling, sequencing, analysis, and reporting, with sequencing strategy being a crucial step offering numerous options. Present mainstream sequencing strategies include Amplicon sequencing, Metagenomic Next-Generation Sequencing (mNGS), and Targeted Next-Generation Sequencing (tNGS). Two innovative technologies recently emerged, namely MobiMicrobe high-throughput microbial single-cell genome sequencing technology and 2bRAD-M simplified metagenomic sequencing technology, compensate for the limitations of mainstream technologies, each boasting unique core strengths. This paper reviews the basic principles and processes of these three mainstream and two novel microbiological technologies, aiding readers in understanding the benefits and drawbacks of different technologies, thereby guiding the selection of the most suitable method for their research endeavours.

FIGURE 1

Schematic Diagram of the Fundamental Procedures in the Sequencing Process Across Five Distinct Microbial Sequencing Technologies. The diagram is segmented into three phases: sample acquisition, sample processing and library construction, and high‐throughput sequencing. After acquiring biological samples, one can either extract nucleic acids directly or process the microbial specimens. Subsequently, various sequencing methods undertake library construction based on their specific protocols. Finally, sequencers are selected for online sequencing, ranging from second to third generation technologies.

FIGURE 2

Structure of amplified target genes and principles of library construction. (A) 16S ribosomal RNA (16S rRNA), component of the 30S subunit of the 70S ribosomal complex in the ribosomes of prokaryotes, is approximately 1542 bp in length, typically comprises 10 conserved regions and 9 variable regions (V1–V9) (Bharti & Grimm, 2019). These conserved and variable regions are arranged alternately. While the conserved regions reflect the affinity among bacterial species, displaying minimal variance, the variable regions denote species‐specific traits, which present appreciable differences across diverse bacteria. The primer design targets the conserved regions and amplifies a variable region (for example, V3) or several variable regions (for example, V3–V4). The amplicon sequencing library is created by introducing adapters and sequencing the resulting amplicons (Schlaberg, 2020). The choice of target areas depends on the sample type and is guided by published articles or experimental testing to determine the most appropriate V‐region (Franzén et al., ; Weinroth et al., 2022). (B) The 18S rDNA encodes the small subunit rRNA of eukaryotic ribosomes and serves as crucial evidence for fungal classification. Similar to 16S rDNA, the 18S rDNA sequence is about 1500–2000 bp long and has both conserved and eight different variable regions (V1–V9, no V6 region) (Banos et al., 2018). For fungi, variable regions V1, V4, V5, and V9 are the most discriminating (Reich & Labes, 2017). Amplicon sequencing based on the 18S rDNA and Internal Transcribed Spacer (ITS) regions is the most common method for molecular identification of fungi. The 18S rDNA genes offer more stability in the fungal community than the ITS region (Liu et al., 2015). (C) Principle of library construction: The steps are divided into two rounds of PCR: during PCR1, the target sequence (blue) is targeted and amplified, the primers contain the index sequence, the heterozygous spacer sequence (red) and part of the Illumina adapter (green); and during PCR2, the introduction of the third index sequence (dark green) as well as the completion of the Illumina sequencing adapter.

FIGURE 3

2bRAD‐M computational workflow. During sequencing, the obtained sequencing results are compared with a comprehensive fingerprint library consisting of 260,000 microbes. For qualitative analysis, only the unique tags are considered, while tags shared by multiple species are excluded. Matching between the sequencing results and the unique tags in the library indicates the presence of specific species in the sample, while non‐matches suggest the absence of those species (as shown in the figure, if there are five microbes with unique tags A, B, C, D, and E in the fingerprint library, and the microbes in the sample match A, B, and D, it means that these three species are present in the sample, and the microbes do not match C and E, it means that these two species are not present in the sample). In addition to qualitative analysis, relative quantitative information of the species is determined. This involves evaluating the sequencing depth of each tag in combination with the number of unique tags. Through the application of an algorithmic formula, relative quantitative information is calculated, providing insights into the abundance of the detected species.

FIGURE 4

Genome Assembly and Strain‐Level Resolution through Comparative Analysis. The MobiMicrobe genome assembly process involves several steps. Firstly, the reads from each Single Amplified Genome (SAG) are assembled into overlapping clusters called contigs. The genomic signature of each SAG is extracted and clustered, grouping together SAGs with similar signatures into the same bin. The reads within each bin are then assembled to generate genomic sequences. The genomic signature information of each bin assembly is extracted, and this iterative process is repeated. Next, the Average Nucleotide Similarity (ANI) is calculated to analyse the similarity between different bin co‐assembled genomes. Reads from bins with an ANI >95% are combined to assemble genomes at the species level. By comparing the SAGs to the species‐level genomes, SNPs (a DNA sequence polymorphism caused primarily by variation in a single nucleotide at the genomic level that can indicate individual specificity) can be identified, enabling the assignment of SAGs to different strains and facilitating the co‐assembly of strain‐specific genomes.