The Genotype-Tissue Expression (GTEx) Pilot Analysis: Multitissue Gene Regulation in Humans
2015 · paper · Biomedicine
Analyzes RNA sequencing data from 1641 samples across 43 tissues from 175 individuals in the GTEx pilot. It describes gene expression across tissues, catalogs tissue-specific and shared expression-regulating variants, and identifies genome-wide association signals explained by them.
4 findings
Substance Text says "A group of 227 samples (from the cerebellum and cortex) forms an independent subcluster (cluster 1), and a smaller group of 97 samples (cluster 2, dominated by the remaining subregions)", but Table 1 lists only 101 cerebellum and cortex samples and 313 brain samples in total.
- p. 3 Gene expression across tissues, 3rd paragraph
group of 227 samples (from the cerebellum and cortex) forms an independent subcluster
- p. 3 Gene expression across tissues, 3rd paragraph
a smaller group of 97 samples (cluster 2, dominated by the remaining
- p. 32 Table 1, brain rows
BRNCHA
Only the authors can say what is right here.
Judge's note: Read the paragraph on p3, Table 1 on p33 and Fig. 1B on p19. Table 1 gives BRNCHA 30, BRNCHB 24, BRNCTXA 23, BRNCTXB 24, which is 101, so 227 samples cannot all be cerebellum and cortex. The 13 brain rows sum to 313, less than 227 + 97 = 324. In Fig. 1B the wider cluster 1 carries a mixed subregion colour bar. I found no statement in the article explaining the counts. The sizes or the descriptions of the two clusters appear swapped or misstated; only the authors can say which.
Second check: another model agreed
Report a problem with this finding
Substance Text: "The number of eGenes ranged from 919 in heart to 2244 in thyroid ... (Fig. 2A)", but the Fig. 2A key prints "Heart [83, 918]" and "Thyroid [105, 2243]".
- p. 4 Single-tissue eQTL analysis, 2nd paragraph
The number of eGenes ranged from 919 in heart to 2244 in thyroid
- p. 20 Fig. 2A legend
Thyroid [105, 2243]
- p. 20 Fig. 2A legend
Heart [83, 918]
- p. 4 Single-tissue eQTL analysis
ranged from 919 in heart to 2244 in thyroid
Only the authors can say what is right here.
Judge's note: Read the p4 text and the Fig. 2A key on the p20 image. The bracketed sample sizes (83, 105) match Table 1, so the second number is the eGene count. Both values differ from the text by one. The article gives no explanation; the difference is small and either source could be the correct one.
Second check: another model agreed
Report a problem with this finding
Reference or label eQTL sharing is cited to Fig. 2A in two places: "with lower sharing of eQTLs and lower average replication in other tissues (Fig. 2A, fig. S12, and figs. S20 to S22)" and "shared at lower levels with other tissues, as was observed for eQTLs (Fig. 2A and fig. S12, A and B)"; the sharing heat map is Fig. 2B.
- p. 8 ASE section, end of the paragraph beginning 'We found that the overall tissue specificity'
with lower sharing of eQTLs and lower average replication in other tissues (Fig. 2A, fig. S12, and figs. S20 to S22)
- p. 9 Analysis of splicing QTLs, 3rd paragraph
as was observed for eQTLs (Fig. 2A and fig. S12, A and B)
- p. 20 Fig. 2 legend
(B) Dendrogram and heat map of pairwise eQTL sharing
- p. 8 ASE tissue sharing paragraph
lower average replication in other tissues (Fig. 2A, fig. S12
Only the authors can say what is right here.
Judge's note: Read p8, p9 and Fig. 2 with its legend on p20. Fig. 2A shows eGene counts against sample size. Fig. 2B is the pairwise sharing heat map, and its legend says "Blood has the lowest levels of eQTL sharing with other tissues". Fig. 2A does show blood and muscle on a shallower slope, so the p8 pointer could be meant loosely, but the p9 statement is specifically about sharing, which is Fig. 2B.
Second check: another model agreed
Report a problem with this finding
Reference or label "The higher variance in eQTL heterozygotes than in homozygotes (Mann-Whitney P = 2.13 × 10−7; Fig. 4B)": the variance comparison with this P value is Fig. 4C; Fig. 4B is the NDRG4 example.
- p. 9 ASE section, last paragraph
(Mann-Whitney P = 2.13 × 10−7; Fig. 4B)
- p. 23 Fig. 4 caption, panel C
heterozygotes show higher variance in allelic ratio (Mann-Whitney P = 2.13 × 10−7). (D)
- p. 23 Fig. 4 legend, panel C
heterozygotes show higher variance in allelic ratio (Mann-Whitney P = 2.13 × 10−7)
- p. 9 ASE/eQTL paragraph
Mann-Whitney P = 2.13 × 10−7; Fig. 4B
Judge's note: Read p9 and Fig. 4 with its legend on p23. The legend assigns the heterozygous-versus-homozygous variance result (Mann-Whitney P = 2.13 × 10−7) to panel C, the variance scatter. Panel B is the NDRG4 allele-ratio bar plot with linear regression P = 5.69 × 10−6. The pointer should be Fig. 4C.
Second check: another model agreed
Report a problem with this finding
12 minor (typos and the like; most checked by the judge only)
Reference or label "Blood and LCLs have a higher proportion of allelic effects (3.6% versus 2.3% in others; P < 2 × 10−16, Fisher’s exact test) (Fig. 3B)": Fig. 3B shows only the nine eQTL tissues and has no LCL; the all-tissue proportions including LCL are in Fig. 3A.
- p. 7 ASE section, 2nd paragraph
(3.6% versus 2.3% in others; P < 2 × 10−16, Fisher’s exact test) (Fig. 3B)
- p. 22 Fig. 3 caption, panel B
Proportion of significant ASE sites for the nine tissues with eQTL data
- p. 22 Fig. 3 legend, panel B
for the nine tissues with eQTL data
- p. 7 ASE results
3.6% versus 2.3% in others; P < 2 × 10−16, Fisher’s exact test) (Fig. 3B)
Only the authors can say what is right here.
Judge's note: Read the p7 text crop and Fig. 3 on p22. Panel A shows per-tissue ASE proportions with LCL and WHLBLD as the highest bars. Panel B plots only the nine eQTL tissues; WHLBLD is the top point and LCL is absent. The pointer is therefore partly relevant (blood) but does not show LCL or the comparison with the other tissues; Fig. 3A, or 3A and 3B, seems intended.
Second check: another model agreed
Report a problem with this finding
Wording "Notably, both methods indicate that more than 50% of all detected eQTLs are common to all nine tissues", while Fig. 2C plots both methods at about log10 P = -0.37 for nine tissues (roughly 43%).
- p. 5 Multitissue joint discovery, 4th paragraph
both methods indicate that more than 50% of all detected eQTLs are common to all nine tissues
- p. 20 Fig. 2C, y-axis value at x = 9
P (eQTL active
- p. 20 Fig. 2C, points at x = 9
P (eQTL active in x tissues)
- p. 5 Tissue specificity and sharing
more than 50% of all detected eQTLs are common to all nine tissues
Only the authors can say what is right here.
Judge's note: Read the p5 text and Fig. 2C on p20. Both points at x = 9 sit a little above the -0.5 tick, at about -0.37 by my reading, where more than 50% would need a value above -0.30. This is a reading off a small plot, so it is approximate. "All detected eQTLs" may also be a different set from the "randomly selected gene or SNP-gene pair" that Fig. 2C plots. The paper does not say, so this is a query on which quantity the 50% refers to.
Second check: another model agreed
Report a problem with this finding
Wording Fig. 5A gives genotype counts "CC (76)", "CG (53)", "GG (10)" for a skeletal-muscle sQTL ("number of individuals below each genotype"), totalling 139, while Table 1 and Fig. 2A give 138 skeletal muscle samples.
- p. 24 Fig. 5A x-axis genotype counts
(76)
- p. 33 Table 1, Muscle - Skeletal row
138
- p. 24 Fig. 5 legend, panel A
number of individuals below each genotype
- p. 20 Fig. 2A in-figure legend
Muscle [138, 1535]
Only the authors can say what is right here.
Judge's note: Viewed page 24 (counts 76, 53, 10 under the box plots; legend says skeletal muscle), page 33 Table 1 (Muscle - Skeletal n = 138) and page 20 Fig. 2A ("Muscle [138, 1535]"). I found no statement in the article text explaining a different muscle sample set for the sQTL analysis; methods are deferred to the supplement (ref. 14), which is not among the files. A one-sample discrepancy only the authors can resolve.
Second check: another model agreed
Report a problem with this finding
Typo "Significance correlations between genotypes and gene expression levels were determined by linear regression": wrong word form ("Significant correlations" or "Significance of correlations").
- p. 4 first paragraph
Significance correlations between genotypes and gene expression levels were determined
Judge's note: Seen as printed on the p4 page image. Wording slip only.
Second check: none (judge only)
Report a problem with this finding
Typo "proxy for capturing regulatory effects in another tissues of interest": number disagreement ("another tissue" or "other tissues").
- p. 8 ASE section, 1st paragraph on page
proxy for capturing regulatory effects in another tissues of interest
- p. 8 ASE section, 1st paragraph, last sentence
in another tissues of interest
- p. 8 first paragraph, last sentence
capturing regulatory effects in another tissues of interest
Judge's note: Seen as printed on the p8 page image.
Second check: none (judge only)
Report a problem with this finding
Typo Tool name capitalised inconsistently in one paragraph: "sQTLSeekeR (36)" (three times) and then "sQTLseekeR" in the rest of the section.
- p. 9 Analysis of splicing QTLs, first paragraph
sQTLSeekeR (36)
- p. 9 same paragraph
sQTLseekeR
Judge's note: Seen on the p9 page image: "sQTLSeekeR" three times in the first paragraph of the section, then "sQTLseekeR" at the end of that paragraph and throughout the next; the Fig. 5 legend also uses the lower-case form. Cosmetic; one form should be used throughout.
Second check: none (judge only)
Report a problem with this finding
Typo "provides unique insights in to the diversity and regulation of gene expression": "in to" should be "into".
- p. 14 Conclusions
unique insights in to the diversity and regulation
Judge's note: Seen as printed on the crop of p14.
Second check: none (judge only)
Report a problem with this finding
Typo Contract number "10X172 (Science Care Inc.)" breaks the pattern of the adjacent "10XS170 (NDRI), 10XS171 (Roswell Park Cancer Institute)".
- p. 15 Acknowledgments
10X172 (Science Care Inc.)
Only the authors can say what is right here.
Judge's note: Rests on outside knowledge: the true contract number for Science Care Inc., which cannot be checked from the files. The page prints "10X172" directly after "10XS170" and "10XS171", so a dropped "S" is plausible from the pattern alone, but only the authors can confirm.
Second check: none (judge only)
Report a problem with this finding
Typo Fig. 2 title prints "Number and sharing of significant ciseQTLs per tissue"; the hyphen is missing ("cis-eQTLs", as in the first line of the same legend).
- p. 20 Fig. 2 title
Number and sharing of significant ciseQTLs per tissue
Judge's note: Viewed page 20 image: the bold title reads "ciseQTLs" while the next line reads "cis-eQTL genes". The run-together form is in the printed page, not only the text layer.
Second check: none (judge only)
Report a problem with this finding
Typo Fig. 3 legend: "(1) different tissues and different individual" should read "different individuals".
- p. 22 Fig. 3 legend
different tissues and different individual
Judge's note: Viewed page 22 image: the legend prints the singular "different individual," while item (3) in the same sentence has "different individuals".
Second check: none (judge only)
Report a problem with this finding
Typo Fig. 6F panel title reads "Expression across 175 individuals for the ZPF57 gene"; the gene is ZFP57 in the legend and text.
- p. 25 Fig. 6F panel title
Expression across 175 individuals for the ZPF57 gene
- p. 26 Fig. 6 caption, panel F
for the ZFP57 gene encoding a KRAB domain transcription factor
Judge's note: Viewed page 25 image: the panel F title prints "ZPF57". The legend on page 26 ("for the ZFP57 gene encoding a KRAB domain transcription factor") and the text (main.txt line 562) use ZFP57. Letter transposition inside the figure.
Second check: none (judge only)
Report a problem with this finding
Typo Fig. 8 legend: "for known SNP associations with the hypertension" has a superfluous "the".
- p. 29 Fig. 8 caption, last line
for known SNP associations with the hypertension
- p. 29 Fig. 8 legend continuation
known SNP associations with the hypertension
Judge's note: Viewed page 29 image: the legend continuation ends "for known SNP associations with the hypertension." Stray article; cosmetic.
Second check: none (judge only)