assay · intermediate · partly written
RNA-seq
RNA sequencing counts short DNA reads made from RNA; everything else, from which gene a read came from to what an abundance means, is inference.
Also called: RNA sequencing, transcriptome sequencing
The short answer
RNA-seq counts reads, not RNA molecules. It reproduces relative change between samples far better than absolute level, and its numbers are shaped by library preparation, alignment and normalisation before any biology is read. 1
Reported, not yet replicated
What is counted and what is inferred more
Counted: short DNA reads (for example 100 bases) after RNA extraction, reverse transcription, fragmentation and PCR amplification. Inferred: which transcript each read came from when sequences are shared, abundance (reads scaled by length and depth), cross-sample scaling (which assumes most genes do not change), and which differences exceed noise. 1234
Reported, not yet replicated
Documented biases more
Random-hexamer priming biases the nucleotide composition at the start of reads (abstract-level read; the rate was not obtained). In a simulation of 12 workflows on 19,654 protein-coding genes with error-free reads, 958 genes (about 4.9%) were vastly under- or over-counted by at least one workflow, enriched for paralogues. Removing duplicate reads by position did not improve accuracy in a three-protocol comparison, because most flagged duplicates were not PCR artefacts. In breast-tumour RNA, ribosomal depletion versus poly(A) selection changed which genes were found: depletion detected about 550 more annotated genes in one study while poly(A) was most sensitive for differential expression. 3245
Reported, not yet replicated
Reproducibility (SEQC consortium) more
Across sites, up to 95% of differentially expressed genes agreed after filtering; about 44,000 known genes (79%) were detected consistently across replicate site pairs; only 59% of genes titrated correctly between mixed samples; some spike-ins were off by up to 10-fold; HiSeq and SOLiD disagreed on whether 5,056 genes (9%) were expressed. The authors state "none of the technologies we tested could provide reliable absolute quantification" and "there is no single gold standard". In a yeast experiment, three replicates found about 20 to 40% of the genes found with 40 or more; the authors recommend at least six replicates, or twelve when fold change does not matter. 16
Reported, not yet replicated
Relation to protein more
RNA-seq is often used as a stand-in for protein; how good that is depends on the axis. See concepts/mrna-protein-correlation. 78
Reported, not yet replicated
Question answered
Which genes are expressed, and which differ between samples, as read counts of RNA fragments. 1
Reported, not yet replicated
Readout physical
Not measured
Setting
Still open here
- What is the rate of hexamer start bias, and the effect of the Salmon and TMM corrections (full texts not opened)?
- Cost and time per sample were not found in a primary source.
Not yet written: Where in cascade, Sample prep, Conditions, Throughput class, Throughput, Time per run, Cost, Engineerable variants, Use cases, Before after, Capital, Emerging, ML role, Data availability.
Linked from
- Long-read RNA sequencing (nanopore, PacBio)
- Mass-spectrometry proteomics
- Nascent-RNA sequencing (GRO-seq, PRO-seq, NET-seq, mNET-seq, TT-seq)
- Ribosome profiling (Ribo-seq)
- Models of models
- mRNA-protein correlation
- Splicing
- Transcription by RNA polymerase II: initiation, pausing, elongation, termination
- What is the splicing error rate per intron in a living human cell?
Sources
- SEQC/MAQC-III Consortium. A comprehensive assessment of RNA-seq accuracy, reproducibility and information content by the Sequencing Quality Control Consortium. Nat Biotechnol (2014). paper; opened 2026-11-17 all records citing this
- Robert C, Watson M. Errors in RNA-Seq quantification affect genes of relevance to human disease. Genome Biol (2015). paper; opened 2026-11-17 all records citing this
- Hansen KD, Brenner SE, Dudoit S. Biases in Illumina transcriptome sequencing caused by random hexamer priming. Nucleic Acids Res (2010). paper; opened 2026-11-17 all records citing this
- Parekh S, Ziegenhain C, Vieth B, et al.. The impact of amplification on differential expression analyses by RNA-seq. Sci Rep (2016). paper; opened 2026-11-17 all records citing this
- Zhao W, He X, Hoadley KA, et al.. Comparison of RNA-Seq by poly (A) capture, ribosomal RNA depletion, and DNA microarray for expression profiling. BMC Genomics (2014). paper; opened 2026-11-17 all records citing this
- Schurch NJ, Schofield P, Gierliński M, et al.. How many biological replicates are needed in an RNA-seq experiment and which differential expression tool should you use?. RNA (2016). paper; opened 2026-11-17 all records citing this
- Franks A, Airoldi E, Slavov N. Post-transcriptional regulation across human tissues. PLoS Comput Biol (2017). paper; opened 2026-11-17 all records citing this
- Csárdi G, Franks A, Choi DS, et al.. Accounting for experimental noise reveals that mRNA levels, amplified by post-transcriptional processes, largely determine steady-state protein levels in yeast. PLoS Genetics (2015). paper; opened 2026-11-17 all records citing this
Where it sits on the map
History of this record (1)
This record has been changed 1 time, on the dates below. The note kept against each change is written for the people who maintain the atlas, in their own shorthand, so it is not shown here; it is in the record itself.
- Status
- partly written
- Last updated
- 2026-11-17
- Record
- content/assays/rna-seq.json