Current concepts, advances, and challenges in deciphering the human microbiota with metatranscriptomics

Exploring the Microbiome in Human Reproductive Tract: High-Throughput Methods for the Taxonomic Characterization of Microorganisms

“…Meta-transcriptomics (also known as meta-RNAseq) refer to the techniques that aim to address the function of the microbiome by measuring the levels of total RNA in any microbial community 93 (see Fig. 1 ).…”

“…This approach is relatively novel in the field of human reproduction, and the vagina microbiota is the most studied niche. 93 94 With this analysis approach, we do not only detect the microbes that are present but also their potential functions. 93 94 95 96 Accordingly, studies have used meta-transcriptomics to investigate multiple diseases, such as vaginosis or preterm pregnancies.…”

“…93 94 With this analysis approach, we do not only detect the microbes that are present but also their potential functions. 93 94 95 96 Accordingly, studies have used meta-transcriptomics to investigate multiple diseases, such as vaginosis or preterm pregnancies. 97 For example, researchers have been able to describe the previously unknown mechanisms of resistance to the first-line treatment in bacterial vaginosis against Gardnerella vaginalis .…”

“…103 104 However, this is just one step of the action protocols that should be carefully implemented. The most popular software for meta-transcriptomics have been recently reviewed elsewhere (e.g., Ojala et al 93 94 summarized dozens of software and several pipelines and databases), also reviewing human reproduction-related studies, and Hempel et al 105 benchmarked 768 data-processing workflows. Here, we can highlight alignment-based solutions such as SortMeRNA, 106 Bowtie, 83 or STAR 107 to discard rRNA or host sequences; Kraken 50 or MEGAN 108 to perform taxonomic profiling; HUMAnN 89 to profile function; metaSPAdes 85 or MEGAHIT 109 to obtain assemblies; and a dynamic landscape of packages for statistical approaches or differential analyses, 110 such as edgeR, 111 DESeq2, 112 LefSe, or metagenomeSeq, 58 which may perform differently, as shown in benchmarks for 16S rRNA gene sequencing.…”

“…Here, we can highlight alignment-based solutions such as SortMeRNA, 106 Bowtie, 83 or STAR 107 to discard rRNA or host sequences; Kraken 50 or MEGAN 108 to perform taxonomic profiling; HUMAnN 89 to profile function; metaSPAdes 85 or MEGAHIT 109 to obtain assemblies; and a dynamic landscape of packages for statistical approaches or differential analyses, 110 such as edgeR, 111 DESeq2, 112 LefSe, or metagenomeSeq, 58 which may perform differently, as shown in benchmarks for 16S rRNA gene sequencing. 113 In any case, we advocate that users should always be cautious when choosing any bioinformatic pipeline for their analyses, paying due attention to the continuous technological and software advances (e.g., the future implementation of amplification-free long-read sequencing technologies, reviewed elsewhere, 93 ) and avoiding common mistakes. 93 103 104 We recommend users to test previous attempts at streamlining the bioinformatic pipelines and providing general solutions for meta-transcriptomics purposes, such as Aguiar-Pulido et al 96 and Knight et al 114 More recently, Terrón-Camero et al 115 employed as test cases endometrium datasets and provided a generalizable pipeline based on Kraken and Bracken.…”