The DOI matches its Crossref record; 214 of the 214 references it cites were found in a registry or an index; 74 of the 146 sources its claims cite were checked for support.
A survey of best practices for RNA-seq data analysis
Open accessAbstract
RNA-sequencing (RNA-seq) has a wide variety of applications, but no single analysis pipeline can be used in all cases. We review all of the major steps in RNA-seq data analysis, including experimental design, quality control, read alignment, quantification of gene and transcript levels, visualization, differential gene expression, alternative splicing, functional analysis, gene fusion detection and eQTL mapping. We highlight the challenges associated with each step. We discuss the analysis of small RNAs and the integration of RNA-seq with other functional genomics techniques. Finally, we discuss the outlook for novel technologies that are changing the state of the art in transcriptomics.
Overview
CoChat Summary
TL;DR
This review surveys best practices for RNA-seq data analysis, covering experimental design, quality control, alignment, quantification, differential expression, splicing, fusion detection, small RNAs, functional profiling, data integration, and emerging single-cell/long-read technologies. It stresses that no single optimal pipeline exists and that analysis choices should depend on organism, annotation quality, and biological question.
Key findings
- No single RNA-seq analysis pipeline is optimal; choices vary by organism, genome/transcriptome availability, and goals, with genome mapping enabling novel transcript discovery and transcriptome mapping being faster but limited to known transcripts.
- Experimental design depends on rRNA removal strategy, strand specificity, read length/paired-end status, and sequencing depth: 5 million mapped reads may suffice for medium-to-highly expressed eukaryotic genes, while up to 100 million reads may be used for low-expression genes; single-cell quantification often uses 1 million reads, with as few as 50,000 or even 20,000 reads used in some cases.
- Quality control expects 70–90% of regular RNA-seq reads to map to the human genome, technical replicate reproducibility of Spearman R² > 0.9, and recommends discarding outliers with over 30% disagreement.
- Differential expression method choice matters: limma performs well and is fast; DESeq and edgeR rank genes similarly but are often conservative or liberal, respectively, in FDR control; SAMseq performs well with at least 10 replicates; Cuffdiff/Cuffdiff2 performed poorly in comparisons. Power calculations in Table 1 show, for example, 87% power at 3 replicates, 98% at 5, and 100% at 10 for a twofold effect size.
Methods
- Design: narrative review/survey of best practices for RNA-seq data analysis, not a primary experimental study; it provides commented guidelines rather than an exhaustive tool compilation or a single recommended pipeline.
- Data/sample: synthesizes published literature, software tools, and resources across major RNA-seq analysis steps; includes an illustrative power calculation using a human whole-blood RNA-seq sample sequenced at 30 million mapped reads from the GTEx Project, with median coverage of 70 aligned reads for a protein-coding gene.
- Main analysis: qualitative synthesis organized by analysis stage, with boxes on replicates, experiment execution, mapping, and differential expression tool comparisons, plus a table of statistical power across effect sizes, sequencing depths, and replicate numbers.
Limitations
- Stated: the review is not exhaustive, does not indicate one best analysis pipeline, and notes that no clear consensus exists on best practices; comparison studies are often based on simulated datasets or spike-in data with minimal variation, and results are affected by parameter settings, especially for lowly expressed genes.
- Stated: short-read sequencing remains limited for accurate full-length transcript reconstruction and isoform-level analysis; long-read technologies still have high error rates and low throughput; single-cell RNA-seq faces technical noise, low capture rates (around 10–50%), and frequent loss of lowly expressed transcripts.
- Evident: as a narrative review, it lacks a systematic search strategy or formal quality appraisal, and its tool recommendations reflect the state of the field in 2016, so performance rankings and software versions may be outdated.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Figures


Verification
Claims the paper makes
35 Valid · 22 Not valid · 45 Not checked · 41 Own results and interpretations
Some cited sources support this claim and some don’t — read each source’s verdict below.
In the paper
“Cuffdiff and Cuffdiff2 have performed surprisingly poorly in the comparisons [56, 63].”
§ A survey of best practices for RNA-seq data analysis > Box 4. Comparison of software tools for detecting differential gene and transcript expression
What the sources say
- Disputedcites [56] Soneson C, Delorenzi M. A comparison of methods for differential expression analysis of RNA-seq data. BMC Bioinformatics. 2013;14:91. sourceExperimental — accuracy not yet measured
Our copy states in Results and discussion (and Background) that exactly eleven methods were compared — DESeq, edgeR, NBPSeq, TSPM, baySeq, EBSeq, NOISeq, SAMseq, ShrinkSeq, voom+limma and vst+limma — and the study was confined to R methods applicable to count matrices. Cufflinks/Cuffdiff appears only once, in the Background (p. 1), as an example of external software whose inference is based on FPKM values, i.e. not among the evaluated methods; Cuffdiff2 is not mentioned anywhere, and Table 2 (Conclusions) summarizes observations for the eleven methods only. Thus the paper cannot be the source of a finding that Cuffdiff/Cuffdiff2 'performed surprisingly poorly in the comparisons' — it says the opposite about their scope. (LLM, conf=0.85)
Checked by a reasoning model — may be wrong.
- Partly supportedcites [63] Rapaport F, Khanin R, Liang Y, Pirun M, Krek A, Zumbo P, et al. Comprehensive evaluation of differential gene expression analysis methods for RNA-seq data. Genome Biol. 2013;14:R95. sourceExperimental — accuracy not yet measured
The Background (p. 1) states that 'In most benchmarks Cuffdiff performed less favorably with a higher number of false positives without any increase in sensitivity', and the null-model section (Results, Figure 3 / Table 1 discussion) and the genes-expressed-in-one-condition section (Figure 4) document Cuffdiff's inflated false positives and reduced specificity — so the poor performance of Cuffdiff in the comparisons is supported. However, our copy evaluates only one Cuffdiff implementation, listed in Materials and methods ('Normalization and differential expression') as 'cuffdiff (v.2.0.0 (3365))'; it never compares Cuffdiff and Cuffdiff2 as two distinct tools, so the claim's attribution of poor performance to both versions is not what this paper reports. The 'surprisingly' framing is also the citing authors' gloss; the paper presents the result plainly (and notes Anders et al. observed the same inflation). (LLM, conf=0.78)
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Disputed
Every source the paper cites for this claim was checked and supports it.
In the paper
“Saturation curves can be used to assess the improvement in transcriptome coverage to be expected at a given sequencing depth [10].”
§ A survey of best practices for RNA-seq data analysis > Experimental design
What the sources say
- Supportedcites [10] Tarazona S, Garcia-Alcalde F, Dopazo J, Ferrer A, Conesa A. Differential expression in RNA-seq: a matter of depth. Genome Res. 2011;21:2213–23. source
The 'Saturation, gene length, and reads distribution' section defines saturation curves as plotting 'the number of genes detected by more than five uniquely mapped reads as a function of the sequencing depth', alongside the new detections rate (NDR, newly detected genes per million additional reads), which directly quantifies incremental gains in transcript detection. The section's concluding paragraph explicitly recommends using the per-biotype transcript detection and length accumulative curves 'to estimate the saturation and contamination levels of any particular mRNA-seq data set', and the introduction frames knowledge of the relationship among SD, feature detection and differential expression as needed 'for experimental design purposes'. Thus our copy supports that saturation curves assess how transcriptome coverage improves with sequencing depth. (LLM, conf=0.88)
plainstates an association- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Supported
Every source the paper cites for this claim was checked and supports it.
In the paper
“When studying single cells, which have limited sample complexity, quantification is often carried out with just one million reads but may be done reliably for highly expressed genes with as few as 50,000 reads [8]; even 20,000 reads have been used to differentiate cell types in splenic tissue [9].”
§ A survey of best practices for RNA-seq data analysis > Experimental design
What the sources say
- Supportedcites [9] Jaitin DA, Kenigsberg E, Keren-Shaul H, Elefant N, Paul F, Zaretsky I, et al. Massively parallel single-cell RNA-seq for marker-free decomposition of tissues into cell types. Science. 2014;343:776–9. source
Our copy (Abstract, p.1) states that 'multiplexing 1536 cells in one sequencing lane provided an average of 22 thousands aligned reads per cell' and that this low-depth sampling was used to classify the splenic cell population into distinct types (class I–V linked to B cells, NK cells, macrophages, monocytes and pDC, remaining classes to DC). The claim's '~20,000 reads to differentiate cell types in splenic tissue' thus rests on the paper's own ~22,000-reads-per-cell average, so the cited figure is a slight downward rounding of the reported depth rather than a number the paper prints; population, depth order of magnitude and the cell-type discrimination are all in our copy. (LLM, conf=0.80)
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Supported
The sources with a verdict support this claim; the rest weren’t checked yet, have no full text we could read, or their verdicts aren’t rated yet.
In the paper
“For example, rRNA and small RNAs should not be present in regular polyA longRNA preparations [10, 19].”
§ A survey of best practices for RNA-seq data analysis > Analysis of the RNA-seq data > Quality-control checkpoints
What the sources say
- Supportedcites [10] Tarazona S, Garcia-Alcalde F, Dopazo J, Ferrer A, Conesa A. Differential expression in RNA-seq: a matter of depth. Genome Res. 2011;21:2213–23. source
In 'Saturation, gene length, and reads distribution' our copy states that 'rRNA depletion usually precedes mRNA preparation, and rRNA presence is considered as a contamination in mRNA-seq experiments' and that 'In general, these small RNA species can be considered as not targeted by the mRNA-seq procedure,' covering both rRNA and small RNAs (tRNAs, snRNAs, snoRNAs, miRNAs). The same section establishes the basis for the 'regular polyA longRNA' qualifier: the datasets followed the standard Illumina mRNA protocol with poly-A isolation and gel size selection, so 'transcripts should be polyadenylated and larger than the size selection cutoff—typically ~200 bp—to be captured.' The Discussion and Fig. 2 results reinforce this, noting small RNAs were only marginally detected and appear as off-target species at ultra-high depth, consistent with the claim's 'should not be present' framing rather than contradicting it. (LLM, conf=0.87)
- Not checked yetcites [19] Tarazona S, Furió-Tarí P, Turrà D, Pietro AD, Nueda MJ, Ferrer A, et al. Data quality aware analysis of differential expression in RNA-seq with NOISeq R/Bioc package. Nucleic Acids Res. 2015;43:e140.
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Supported
Every source the paper cites for this claim was checked and supports it.
In the paper
“In addition, it is difficult to identify transcription start and end sites [21], and tools such as GRIT [22] that incorporate other data such as 5’ ends from CAGE or RAMPAGE typically have a better chance of annotating the major expressed isoforms correctly.”
§ A survey of best practices for RNA-seq data analysis > Analysis of the RNA-seq data > Transcript identification
What the sources say
- Supportedcites [22] Boley N, Stoiber MH, Booth BW, Wan KH, Hoskins RA, Bickel PJ, et al. Genome-guided transcript assembly by integrative analysis of RNA sequence data. Nat Biotechnol. 2014;32:341–6. source
The Discussion states that GRIT 'leverages multiple RNA sequence data types, including CAGE, mRNA-seq, polyA+site-seq, ESTs and cDNAs to discover transcript models', with models beginning at a transcript start site, and that it 'performs substantially better than competing methods, both at identifying previously annotated transcript models and discovering new genes and transcripts'. The Results > Comparison to other tools quantifies this (GRIT recovers 44.2% of FlyBase transcripts at 17.8% precision, 39.8% at 41.3% after filtering, versus ≤13.4% recall for Cufflinks, Trinity+Rsem and Scripture; GRIT expression scores correlate 0.71–0.80 with CAGE and poly(A)-site-seq tags versus <0.5 for the others), and the Abstract frames this as underscoring 'the importance of integrating multiple types of assays'. Only the illustrative assay name 'RAMPAGE' does not appear in our copy (the paper's 5'-end data are CAGE and RACE), but the attributed point — that incorporating 5'-end/gene-boundary assays such as CAGE gives a better chance of correctly annotating the expressed isoforms — is stated by the source. (LLM, conf=0.82)
hedgedstates a possible cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Supported
57 more claims · 23 with their cited sources checked
Create a free account to see all 143 claims and to verify the 72 references still open.
- It may state two separate claims as one. 52
- Probably not a substantial claim. 35
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
The source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“To identify differentially expressed genes between the cases and controls, the power provided by more samples is generally preferable to the increased accuracy provided by higher depth of sequencing.”
From Multi-omics approaches to disease (2017) · cites this paper as [8]
Experimental — accuracy not yet measuredIn the citing paper
“However, if the main purpose of the analysis is to identify new transcripts, or examine allele-specific expression, the higher depth of coverage is desirable [7–9]”
From Multi-omics approaches to disease (2017) · cites this paper as [8]
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
Each contradiction between two passages listed here was confirmed by an independent check.
Rated minor: it would not touch a result — a value no result rests on, a p-value that stays on the same side of the significance level, or a conflict that is clerical, resolved elsewhere in the paper, or outside its own findings.
In the paper
§ A survey of best practices for RNA-seq data analysis > Background
“quantification would be achieved by first assembling reads de novo into contigs and then mapping these contigs onto the transcriptome.”
§ A survey of best practices for RNA-seq data analysis > Analysis of the RNA-seq data > Transcript identification
“For quantification, reads are mapped back to the novel reference transcriptome”
Rated minor: it would not touch a result — a value no result rests on, a p-value that stays on the same side of the significance level, or a conflict that is clerical, resolved elsewhere in the paper, or outside its own findings.
In the paper
§ A survey of best practices for RNA-seq data analysis > Analysis of the RNA-seq data > Transcript quantification
“Algorithms that quantify expression from transcriptome mappings include RSEM (RNA-Seq by Expectation Maximization) [40], eXpress [41], Sailfish [35] and kallisto [42] among others.”
§ A survey of best practices for RNA-seq data analysis > References
“34. Patro R Mount SM Kingsford C Sailfish enables alignment-free isoform quantification from RNA-seq reads using lightweight algorithms Nat Biotechnol 2014 32 462 4”
Rated minor: it would not touch a result — a value no result rests on, a p-value that stays on the same side of the significance level, or a conflict that is clerical, resolved elsewhere in the paper, or outside its own findings.
In the paper
§ A survey of best practices for RNA-seq data analysis > Analysis of the RNA-seq data > Transcript identification
“using packages such as SOAPdenovo-Trans [30], Oases [31], Trans-ABySS [32] or Trinity [33].”
§ A survey of best practices for RNA-seq data analysis > References
“32. Grabherr MG Haas BJ Yassour M Levin JZ Thompson DA Amit I Full-length transcriptome assembly from RNA-seq data without a reference genome Nat Biotechnol 2011 29 644 52”
Rated minor: it would not touch a result — a value no result rests on, a p-value that stays on the same side of the significance level, or a conflict that is clerical, resolved elsewhere in the paper, or outside its own findings.
In the paper
§ A survey of best practices for RNA-seq data analysis > Analysis of the RNA-seq data > Alternative splicing analysis
“as discussed in the RNA-seq Genome Annotation Assessment Project paper [30].”
§ A survey of best practices for RNA-seq data analysis > References
“30. Xie Y Wu G Tang J Luo R Patterson J Liu S SOAPdenovo-Trans: de novo transcriptome assembly with short RNA-Seq reads Bioinformatics 2014 30 1660 6”
Rated minor: it would not touch a result — a value no result rests on, a p-value that stays on the same side of the significance level, or a conflict that is clerical, resolved elsewhere in the paper, or outside its own findings.
In the paper
§ A survey of best practices for RNA-seq data analysis > Analysis of the RNA-seq data > Differential gene expression analysis
“the GC contents of genes (corrected in the EDAseq package [21]).”
§ A survey of best practices for RNA-seq data analysis > Analysis of the RNA-seq data > Quality-control checkpoints
“A number of R packages (such as NOISeq [19] or EDASeq [20]) provide useful plots for quality control of count data.”
2 more findings
Create a free account to see every finding of this check.
Citations
Cites
Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2
Love, M. I., Huber, W., & Anders, S. (2014) · Genome Biology · doi:10.1186/s13059-014-0550-8 · Cited by 95,106 · cited 2×
“DESeq2, like edgeR, uses the negative binomial as the reference distribution and provides its own normalization approach [48, 58]. baySeq [59] and EBSeq [60] are Bayesian approaches, also based on the negative binomial model, that define a collection of models to describe the differences among…”
Trimmomatic: a flexible trimmer for Illumina sequence data
Bolger, A. M., Lohse, M., & Usadel, B. (2014) · Bioinformatics · doi:10.1093/bioinformatics/btu170 · Cited by 65,548 · cited 1×
“Software tools such as the FASTX-Toolkit [13] and Trimmomatic [14] can be used to discard low-quality reads, trim adaptor sequences, and eliminate poor-quality bases.”
MethodFast and accurate short read alignment with Burrows-Wheeler transform
Li, H., & Durbin, R. (2009) · Bioinformatics · doi:10.1093/bioinformatics/btp324 · Cited by 58,607
Fast gapped-read alignment with Bowtie 2
Langmead, B., & Salzberg, S. L. (2012) · Nature Methods · doi:10.1038/nmeth.1923 · Cited by 56,400 · cited 1×
“As in standard RNA-seq, sRNA reads must then be aligned to a reference genome or transcriptome sequences using standard tools, such as Bowtie2 [97], STAR [15], or Burrows-Wheeler Aligner (BWA) [98].”
MethodSTAR: ultrafast universal RNA-seq aligner
Dobin, A., Davis, C. A., Schlesinger, F., Drenkow, J., Zaleski, C., Jha, S., Batut, P., Chaisson, M., & Gingeras, T. R. (2013) · Bioinformatics · doi:10.1093/bioinformatics/bts635 · Cited by 51,709 · cited 3×
“Several other mappers exist that are optimized to identify SNPs or indels (GSNAP [202], PALMapper [203] MapSplice [204]), detect non-canonical splice junctions (STAR [15], MapSplice [204]), achieve ultra-fast mapping (GEM [205]) or map long-reads (STAR [15]).”
BackgroundMethodGene Ontology: tool for the unification of biology
Ashburner, M., Ball, C. A., Blake, J. A., Botstein, D., Butler, H., Cherry, J. M., Davis, A. P., Dolinski, K., Dwight, S. S., Eppig, J. T., et al. (2000) · Nature Genetics · doi:10.1038/75556 · Cited by 43,206 · cited 1×
“Resources such as Gene Ontology [109], Bioconductor [110], DAVID [111, 112] or Babelomics [113] contain annotation data for most model species.”
BackgroundRobinson, M. D., McCarthy, D. J., & Smyth, G. K. (2010) · Bioinformatics · doi:10.1093/bioinformatics/btp616 · Cited by 41,364 · cited 2×
“While some of the differential expression tools can only perform a pair-wise comparison, others such as edgeR [57], limma-voom [55], DESeq [48], DESeq2 [58], and maSigPro [213] can perform multiple comparisons, include different covariates or analyze time-series data. of biological interest to…”
MethodSystematic and integrative analysis of large gene lists using DAVID bioinformatics resources
Huang, D. W., Sherman, B. T., & Lempicki, R. A. (2009) · Nature Protocols · doi:10.1038/nprot.2008.211 · Cited by 35,890
featureCounts: an efficient general purpose program for assigning sequence reads to genomic features
Liao, Y., Smyth, G. K., & Shi, W. (2014) · Bioinformatics · doi:10.1093/bioinformatics/btt656 · Cited by 27,219 · cited 1×
“The simplest approach to quantification is to aggregate raw counts of mapped reads using programs such as HTSeq-count [35] or featureCounts [36].”
MethodRSEM: accurate transcript quantification from RNA-Seq data with or without a reference genome
Li, B., & Dewey, C. N. (2011) · BMC Bioinformatics · doi:10.1186/1471-2105-12-323 · Cited by 22,179 · cited 2×
“Additionally, the RSEM algo-rithm uses an expectation maximization approach that returns TPM values [40].”
Method
Cited by
Integrating single-cell transcriptomic data across different conditions, technologies, and species
Butler, A., Hoffman, P., Smibert, P., Papalexi, E., & Satija, R. (2018) · Nature Biotechnology · doi:10.1038/nbt.4096 · Cited by 13,174
“We next compared Seurat’s performance to widely used batch correction tools that have been applied to both bulk 47 and single-cell genomics data 5 .”
MethodMulti-omics approaches to disease
Hasin, Y., Seldin, M., & Lusis, A. (2017) · Genome Biology · doi:10.1186/s13059-017-1215-1 · Cited by 2,953
Reimand, J., Isserlin, R., Voisin, V., Kucera, M., Tannus-Lopes, C., Rostamianfar, A., Wadi, L., Meyer, M., Wong, J., Xu, C., et al. (2019) · Nature Protocols · doi:10.1038/s41596-018-0103-9 · Cited by 1,970
RNA sequencing: the teenage years
Stark, R., Grzelak, M., & Hadfield, J. (2019) · Nature Reviews Genetics · doi:10.1038/s41576-019-0150-2 · Cited by 1,915
Differential analysis of RNA-seq incorporating quantification uncertainty
Pimentel, H., Bray, N. L., Puente, S., Melsted, P., & Pachter, L. (2017) · Nature Methods · doi:10.1038/nmeth.4324 · Cited by 1,534
Metadata
- Authors
- Conesa, A., Madrigal, P., Tarazona, S., Gomez-Cabrero, D., Cervera, A., McPherson, A., Szcześniak, M. W., Gaffney, D. J., Elo, L. L., Zhang, X., & Mortazavi, A.
- Affiliations
- University of Florida (US)
- Wellcome/MRC Cambridge Stem Cell Institute (GB)
- Centro de Investigacion Principe Felipe (ES)
- University of Helsinki (FI)
- Simon Fraser University (CA)
- Adam Mickiewicz University in Poznań (PL)
- Wellcome Sanger Institute (GB)
- Åbo Akademi University (FI)
- Tsinghua University (CN)
- University of California, Irvine (US)
- Corresponding author
- Conesa, Ana
- Mortazavi, Ali
- ORCID
- Venue
- Genome Biology
- Published
- 2016-01-26
- Volume
- 17
- Issue
- 1
- Pages
- 13-13
- ISSN
- 1465-6906 (ISSN-L)
- 1474-760X
- PMC
- PMC4728800
- OpenAlex
- W2236822143
- Semantic Scholar
- 36e00caa724a4f8865b6f675fa27f7297e51ba41
- Cited by
- 2,794
- References
- 214
- Citations by year
- 2026: 238 · 2025: 264 · 2024: 287 · 2023: 284 · 2022: 285 · 2021: 301 · 2020: 286 · 2019: 296 · 2018: 290 · 2017: 220 · 2016: 40
- FWCI
- 25.20
- Type
- journal article
- Language
- en
- Update notices
- correction (2016-08-26)
- Open access
- gold
- doi.org
- Licence
- cc-by
- creativecommons.org/licenses/by/4.0/ (tdm)
- Locations
- Genome biology (publishedVersion, cc-by)
- PubMed Central (publishedVersion, cc-by)
- PubMed (publishedVersion)
- eScholarship (California Digital Library) (submittedVersion, cc-by)
- Työväentutkimus Vuosikirja (submittedVersion, cc-by)
- University of Florida Digital Collections (University of Florida) (submittedVersion, other-oa)
- RiuNet (Universitat Politècnica de València) (submittedVersion, cc-by-nc-nd)
- Apollo (University of Cambridge) (submittedVersion, cc-by)
- Funders
- Seventh Framework Programme (36000 - STATegra)
- National Basic Research Program of China (2012CB316504)
- JDRF (2-2013-32)
- Sigrid Juséliuksen Säätiö
- Topics
- Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Genomics and Phylogenetic Studies
- Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Single-cell and spatial transcriptomics
- Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Gene expression and cancer classification
- Keywords
- RNA-seq data analysis, best practices, experimental design, quality control, read mapping, gene expression quantification, transcript quantification, differential gene expression, alternative splicing, fusion gene detection, eQTL mapping, transcriptomics
- MeSH terms
- Base Sequence, RNA, Software, Alternative Splicing, Sequence Analysis, RNA, Gene Expression Profiling, Genomics, Gene Fusion, High-Throughput Nucleotide Sequencing
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
The claim verdicts this report counts are experimental: their accuracy is not yet measured.
A language model drew the claims from the paper on 2026-10-05, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).
A second model labelled each claim on 2026-10-05 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).
Cited sources are checked with a language model’s help (qwen3.8-omni-flash).
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
How to cite
Cite CoChat for the checks on this page, and the paper for its own findings.
- This page
- CoChat. (2026). Citation check: A survey of best practices for RNA-seq data analysis. https://app.cochat.ai/papers/10.1186/s13059-016-0881-8
- The paper
- Conesa, A., Madrigal, P., Tarazona, S., Gomez-Cabrero, D., Cervera, A., McPherson, A., Szcześniak, M. W., Gaffney, D. J., Elo, L. L., Zhang, X., & Mortazavi, A. (2016). A survey of best practices for RNA-seq data analysis. Genome Biology. https://doi.org/10.1186/s13059-016-0881-8
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.