Introduction
Small RNA sequencing is a widely used technique to study microRNA(miRNAs), small interfering RNAs (siRNAs), piwi-interacting RNA (piRNAs) that play a crucial role in RNA-mediated gene silencing process or known as RNA silencing /Gene silencing process. RNA silencing process employs different types of substrates which give rise to different types of RNA population, namely microRNAs, siRNAs, etc. In the laboratory, small RNA sequencing typically start by extraction of RNA from cells or tissues, followed by Adapter ligation to the 5' and 3' ends of small RNAs, followed by Reverse transcription and PCR amplification to generate cDNA libraries. Finally, High-throughput sequencing ( most commonly Illumina platforms) is used to produce millions of short reads. These resulting data then undergo computational processing to align reads to reference genomes of particular species or miRNA databases.
Bioconductor in RNA Biology
Bioconductor(BioC) [2] is a widely used open-source platform for analysing different types of small-RNA sequencing and genomic data. It primarily utilizes the R programming language and offers a wide range of packages for bioinformatics and computational biology. Bioconductor provides a wide range of packages [3] for handling small-RNA seq data among them few are widely used by researchers. Popular Bioconductor packages like DESeq2, [4] edgeR, [5] limma + voom, [6] [7] GenomicAlignment, [8] GenomicFeatures, [8] Rsubread, [9] ShortRead, [10] featureCounts [11] provide robust analysis of RNA-seq data. [12]
It uses a negative binomial distribution modeling for differential expression analysis of read count from RNA-seq data. [13] It is popular for dispersion estimation, normalization, and visualization by PCA plots or MA plots. [4]
edgeR
It also uses a negative binomial distribution modeling for differential expression analysis of read count from RNA seq data. In contrast with DESeq2, it is used when sample number is relatively small. [5] [13]
limma + voom
It is used to estimate the mean-variance relationship of count data and transforms it to log2-counts per million (CPM). It is used for analysing microarray data and also to calculate CPM value from RNA-seq data. [6]
GenomicAlignment
It is widely popular for reads like BAM and SAM file alignment to assign aligned reads to genes or miRNAs for downstream analysis. [8] [14]
GenomicFeatures
It is used to build transcript-centric annotation databases like TxDb objects which store information about genes, exons, transcripts from GTF/GFF files. [8] [15]
Rsubread
It is used mostly for summarizing the reads and mapping, where functions like align(), featureCounts() are used to provide an efficient alternative to external aligners like STAR or HISAT2. [16]
Computational Workflow
Data Import and Quality Control
FASTQ files [17] are typically imported by using different Bioconductor packages like ShortRead [10] which provides quality assessment reports.
Adapter Trimming and Filtering
Different external tools like Cutadapt, [18] trimmomatic [19] is used to remove the adapter sequence from the raw FASTQ files. This helps to improve the Reads quality.
Read Alignment
The processed Reads are aligned to the reference genome. This alignment can be done by different aligners like Rsubread, or external tools such as STAR, with results stored in standard formats like SAM (Sequence Alignment Map) or BAM (Binary Alignment Map) files .
Annotation of microRNAs
Bioconductor supports to integrate miRBase data where different packages like miRBaseConverter, [20] AnnotationHub, [21] org.Mm.eg.db [22] are used for annotate reads to known miRNAs.
Quantification
Count reads are mapped to known genes or microRNAs, and summarize counts across samples.
Differential Expression Analysis
After mapping and quantifying microRNA expression, different well-established packages like DESeq2, edgeR are used for differential expression analysis.
