Overview
RNA sequencing (RNA-seq) has become the standard method for measuring gene expression, replacing earlier hybridization-based approaches. By converting RNA molecules into a cDNA library and sequencing millions of fragments, RNA-seq provides both the identity and abundance of transcripts in a biological sample. Unlike microarrays, RNA-seq can detect novel transcripts, splice variants, and non-coding RNAs without prior probe design. The technique delivers a dynamic range spanning several orders of magnitude and single-base resolution. RNA-seq data underpin most modern transcriptomics studies, from model organisms to clinical specimens.
Key Concepts
The RNA-seq analysis pipeline begins with raw sequencing reads, which undergo quality control (FastQC), adapter trimming (Trimmomatic or Cutadapt), and alignment to a reference genome using splice-aware aligners such as STAR or HISAT2. Quantification at the gene or transcript level is performed by tools like featureCounts, RSEM, or Salmon. Expression values are normalized as FPKM, RPKM, or TPM to account for sequencing depth and transcript length. Detection of alternative splicing and novel transcripts requires specialized tools such as Cufflinks or StringTie, which assemble transcripts from aligned reads. Quality metrics include mapping rates, read distribution across gene features, and saturation analysis.
Applications
RNA-seq is applied across virtually all areas of biology. It profiles gene expression changes in disease versus healthy tissue, identifies biomarkers for cancer subtypes, and tracks developmental gene expression programs. In microbiology, RNA-seq reveals transcriptional responses to antibiotics and environmental stress. The technique builds directly on RNA sequencing methods and is frequently integrated with next-generation sequencing workflows. RNA-seq also enables discovery of novel RNA structures and types, including long non-coding RNAs and circular RNAs, expanding our understanding of the transcriptome’s complexity.
Practical Protocol
A standard bulk RNA-seq pipeline begins with raw FASTQ files from the sequencer. First, run FastQC to assess quality metrics including per-base quality scores, GC content, adapter contamination, and overrepresented sequences. Trim low-quality bases and residual adapters with Trimmomatic or Cutadapt: a typical command is trimmomatic PE sample_R1.fastq sample_R2.fastq -baseout trimmed.fq ILLUMINACLIP:adapters.fa:2:30:10 LEADING:3 TRAILING:3 SLIDINGWINDOW:4:15 MINLEN:36. For splice-aware alignment, STAR requires a genome index built with STAR --runMode genomeGenerate --genomeFastaFiles genome.fa --sjdbGTFfile annotation.gtf. Align reads with STAR --runThreadN 8 --genomeDir star_index --readFilesIn trimmed_R1.fastq trimmed_R2.fastq --outSAMtype BAM SortedByCoordinate. Quantification at the gene level uses featureCounts -a annotation.gtf -o counts.txt -t exon -g gene_id aligned.bam, producing a count matrix where rows are genes and columns are samples. For transcript-level quantification, Salmon provides alignment-free estimation: salmon quant -i salmon_index -l A -1 R1.fastq -2 R2.fastq -o quant_output. Quality control metrics to report include uniquely mapped read percentage (ideally >80%), mapping rate to exonic regions, 3’ bias, and rRNA contamination level. Assess library complexity with preseq to estimate whether deeper sequencing would yield additional detected genes. The final count matrix is loaded into R for differential expression analysis with DESeq2 or edgeR. For alignment-free quantification, Salmon’s quasi-mapping approach bypasses the alignment step entirely, providing faster processing while maintaining accuracy: salmon quant -i transcript_index -l A -1 reads_R1.fastq -2 reads_R2.fastq --validateMappings -o quant. The --validateMappings flag improves sensitivity for novel transcripts.
