Analysis of Protein-Coding Genetic Variation in 60,706 Humans
2016 · paper · Biomedicine
Aggregates and analyzes exome sequence data from 60,706 people in the Exome Aggregation Consortium. It reports about one variant every eight exome bases, identifies 3,230 genes strongly depleted of truncating variants, and shows use for filtering candidate disease variants.
8 findings
Substance Text: "Nine variants had sufficient data to support disease association"; in Figure 4d the 'disease associated' bar is about 7 variants (six genes listed), while the adjacent 'benign trait' bar is about 9 (nine genes listed).
- p. 6 ExAC improves variant interpretation, para 3
Nine variants had sufficient data
- p. 31 Figure 4d, 'disease associated' bar
disease
- p. 6 ExAC improves variant interpretation, paragraph 3
Nine variants had sufficient data to support disease association
- p. 8 Figure 4d, bar 'disease associated'
disease associated
Only the authors can say what is right here.
Judge's note: Measured on a 3x crop of Figure 4d (page 31): the 0-25 axis interval spans about 135 px; the 'disease associated' bar is about 37 px high (about 7) and the 'benign trait' bar about 48 px (about 9). The disease-associated bar is plainly lower than its neighbour. Six gene names sit above it (BTD, FLG, G6PD, GNE, PRSS1, SERPINA1) and nine above the benign-trait bar. The text's "nine" matches the benign-trait bar, not the disease-associated one. A gene can carry more than one variant, so six genes with seven variants is self-consistent in the figure; it is the text count that disagrees. Either the text or the figure is wrong; Supplementary Section 5.3 (not in the files) would settle which.
Second check: another model agreed
Report a problem with this finding
Substance "we found 179,774 high-confidence PTVs ... 121,309 of which are singletons" and "out of 58,435 PTVs with an allele count greater than one": 179,774 − 121,309 = 58,465, not 58,435.
- p. 7 Impact of rare protein-truncating variants, para 2
we found 179,774 high-confidence PTVs
- p. 7 Impact of rare protein-truncating variants, para 2
121,309 of which are singletons
- p. 7 Impact of rare protein-truncating variants, para 2
out of 58,435 PTVs with an allele count greater than one
Only the authors can say what is right here.
Judge's note: Digits confirmed on the page 7 image. Singletons are defined on page 2 as "variants seen only once in the data set", so the PTVs with allele count greater than one should be the total minus the singletons, 58,465; the printed figure is 30 lower. Searched the article text for a restriction on the second count (a different filter, an allele count of zero after adjustment, a different variant set) and found none; the PTV definition is deferred to Supplementary Section 6, which is not in the files. At least one of the three numbers is misprinted (65 vs 35 looks like a digit slip), but which one only the authors can say.
Second check: another model agreed
Report a problem with this finding
Reference or label "Deleterious variants are expected to have lower allele frequencies than neutral ones ... demonstrated previously in human population sequencing data and here (Figure 1d, Figure 1e)": panel 1e shows the number of indels by size, not allele frequency; the frequency panel (proportion singleton, frameshift vs in-frame) is 1f.
- p. 4 Inferring variant deleteriousness and gene constraint, para 1
and here (Figure 1d, Figure 1e)
- p. 26 Figure 1 legend
and frequency distribution (proportion singleton; f) of indels, by size
- p. 25 Figure 1 legend
e-f) The number (e)
- p. 3 Main text
(Figure 1D)
Only the authors can say what is right here.
Judge's note: Opened Figure 1 (page 25) and its legend (pages 25-26). Panel d is the proportion of possible variants observed by context and class, panel e is "# of variants" by indel length, panel f is "proportion singletons". The legend says "The number (e) and frequency distribution (proportion singleton; f) of indels". Page 3 itself cites 1F for frameshifts being more often singletons. So the pointer to 1e does not support an allele-frequency statement and 1f was probably intended ("Figure 1d, Figure 1f" or "1e-f"); only the authors can say. Panel d is indirect evidence of selection and is defensible. The capitalisation point raised in the same cluster is covered by C003.
Second check: another model agreed
Report a problem with this finding
Reference or label Figure 2 legend, panel b, says the number of variants is shown "as a function of number of individuals (down-sampled from ExAC)" with saturation "at ~20,000 individuals", but the panel's x-axis is labelled "sampled number of chromosomes".
- p. 27 Figure 2 legend, panel b
as a function of number of individuals (down-sampled from
- p. 27 Figure 2b x-axis
sampled number of chromosomes
- p. 27 Figure 2 legend, panel b
~20,000 individuals
- p. 27 Figure 2 legend, panel b
as a function of number of individuals (down-sampled from ExAC). CpG transitions, the most likely mutational event, begin reaching saturation at ~20,000 individuals
Judge's note: On the p27 image the panel b x-axis is labelled 'sampled number of chromosomes', on a log scale with ticks at 500, 2000, 10000 and 50000 and the curves running beyond 50000 (consistent with about 121,000 chromosomes for 60,706 people). The legend on the same page and the main text on p3 ('beginning at around 20,000 individuals (Figure 2b)') give the unit as individuals. Searched the text for a statement converting between the two; there is none. The CpG curve flattens at roughly 40,000-50,000 on the axis, so the 20,000-individual statement is consistent once doubled; the defect is the unit mismatch between legend and axis, not the value.
Second check: another model agreed
Report a problem with this finding
Reference or label Figure 3 legend, panel f: "MAPS (Figure 2d) is shown for each functional category" - MAPS is Figure 2e; Figure 2d is the doubleton cross-population panel.
- p. 30 Figure 3 legend, panel f
f) MAPS (Figure 2d) is shown for each functional category
- p. 28 Figure 2 legend, panel e
e) The mutability-adjusted proportion of singletons (MAPS) is shown across functional classes
- p. 27 Figure 2 legend, panel d
d) For doubletons (variants with an allele count of 2), mutation rate is positively correlated with
- p. 4 Inferring variant deleteriousness, paragraph 1
deleterious (Figure 2e)
Judge's note: Opened Figure 2 on the p27 image: panel d has y-axis 'proportion cross-population' against 'mutability'; panel e has y-axis 'MAPS'. The Figure 2 legend (p27-28) describes d as doubletons and 'e) The mutability-adjusted proportion of singletons (MAPS)', and the main text on p4 cites 'Figure 2e' for MAPS. The Figure 3 legend on the p30 image prints '(Figure 2d)'. The pointer should be Figure 2e.
Second check: another model agreed
Report a problem with this finding
Reference or label Extended Data Figure 5 legend gives a five-colour key, "Latino (red), African (purple), European (blue), South Asian (yellow) and East Asian (green)", but panel b plots six density curves (an unnamed dark navy curve in addition to a lighter blue one), and the curve that should be South Asian is drawn in orange, not yellow.
- p. 16 Extended Data Figure 5 legend
Latino (red), African (purple), European (blue), South Asian (yellow) and East Asian (green)
- p. 16 Extended Data Figure 5 legend
Populations are their respective colors: Latino (red), African (purple), European (blue), South Asian (yellow) and East Asian (green).
- p. 16 Extended Data Figure 5b density curves
HetHom Ratio
Judge's note: Read the full image of PDF page 16. Each of the four density plots in panel b has six curves: red, purple, green, orange, light blue and dark navy. There is no yellow curve and no in-figure key. The legend names five populations, so one curve is unidentified. That the navy curve is Finnish is an inference from the paper's other figures and is not stated on this page; the mismatch between five named colours and six drawn curves, and yellow versus orange, is on the page regardless.
Second check: another model agreed
Report a problem with this finding
Reference or label Several reference entries are incomplete or garbled: ref 11 "Nat. Genet. 2014" (no volume or pages); ref 29 "Sci. Transl. Med. 2016; 8:322ra9–322ra9" (article number repeated as a range); ref 30 "Am. J. Hum. Genet. 2015:1–17" (no volume); ref 34 "Science (80-.). 2016; 8624:1–8" (reference-manager artefact in the journal name, placeholder volume and pages); ref 35 "bioRxiv. 2015" (no identifier); ref 36 "Nat. Neurosci. 2015; 18" (no pages).
- p. 24 Reference 34
Science (80-.). 2016; 8624:1–8.
- p. 23 Ref 11
Nat. Genet. 2014
- p. 24 Ref 36
Nat. Neurosci. 2015; 18
- p. 24 Ref 29
322ra9–322ra9
Judge's note: Read every entry on the p23 and p24 images; all are printed as quoted. Refs 11, 29, 30, 34 and 36 also lack the [PubMed: ...] tag that most complete entries carry. The missing fields and the '(80-.)' artefact are visible on the page and need no outside fact; the correct volumes and pages would have to be looked up, and the claim that '8624:1-8' is not a real Science volume and page range is outside knowledge that I did not rely on beyond noting the entry is malformed next to the other Science entries (e.g. '2012; 335:823-828'). Ref 35 as a preprint may legitimately have no volume, so it is the weakest item in the set.
Second check: another model agreed
Report a problem with this finding
Wording Text: "We manually curated the evidence of pathogenicity for 192 previously reported pathogenic variants" and "163 variants met American College of Medical Genetics criteria for reclassification as benign or likely benign (Figure 4d)"; the bars of Figure 4d total about 160 variants, of which about 124 are benign or likely benign.
- p. 6 ExAC improves variant interpretation, para 3
for 192 previously reported pathogenic
- p. 6 ExAC improves variant interpretation, para 3
163 variants met American College of Medical Genetics
- p. 29 Figure 4d
number of variants
Only the authors can say what is right here.
Judge's note: Measured Figure 4d (page 31) on a 3x crop. Insufficient evidence about 117 (benign about 67, likely benign about 30, uncertain about 20); classification error about 27 (benign about 25, likely benign about 2); benign trait about 9; disease associated about 7. Total about 160 against 192, and benign plus likely benign about 124 against 163. The text names a category, "no claim of pathogenicity", that has no bar in the figure, and the shortfall of about 32 may be that category; but neither the legend (page 32) nor the text says the figure leaves it out, and adding 32 to 124 gives 156, still not 163. Bar readings are uncertain by a few units, not by 30-40. The detail is in Supplementary Section 5.3 and Table 20, which are not in the files, so only the authors can say how the figure and text reconcile.
Second check: another model agreed
Report a problem with this finding
23 minor (typos and the like; most checked by the judge only)
Wording Methods, Sample filtering: "60,706 samples remained, consisting of ~77% of Agilent (33 Mb target) and ~12% of Illumina (37.7 Mb target) exome captures" accounts for only about 89% of the samples; the remaining ~11% is not described.
- p. 10
consisting of ~77% of Agilent (33 Mb target) and ~12% of Illumina (37.7 Mb target)
Only the authors can say what is right here.
Judge's note: Read the crop of PDF page 10; the sentence is printed as quoted. The two shares sum to 89%. Searched the article text for Agilent and Illumina: no other statement gives the capture kit of the remaining samples. The sentence points to Supplementary Information Section 1.7, which is not among the files, so other capture kits may legitimately make up the rest. The reviewer's suggestion that the Illumina share should be ~23% is a guess with no support in the article. Only the authors or the supplement can say whether a number is wrong or a third category is unstated.
Second check: another model agreed
Report a problem with this finding
Typo Author Information prints "P.F.S is a scientific advisor to Pfizer." with the final period of the initials missing (the contributions list just above prints "P.F.S.,").
- p. 1 Author Information
P.F.S is a scientific advisor to Pfizer.
Judge's note: Checked the crop of PDF page 1. The Author Contributions paragraph prints "P.F.S.," with three periods; the Author Information line prints "P.F.S is". Cosmetic.
Second check: none (judge only)
Report a problem with this finding
Typo Panel letters are capitalised in four pointers on PDF page 3, "(Figure 1D)", "(Figure 1E)", "(Figure 1F)" and "(CpG transitions; Figure 2C)", while the rest of the text and the figures use lowercase (e.g. "(Figure 2c)" four lines earlier, "(Figure 1d, Figure 1e)" on page 4).
- p. 3 first paragraph
(Figure 1D)
- p. 3 first paragraph
(Figure 1E)
- p. 3 first paragraph
(Figure 1F)
- p. 3 last paragraph
(CpG transitions; Figure 2C)
Judge's note: Read the full page 3 image: 1D, 1E, 1F and 2C are capitals; "Figure 2a", "Figure 2b", "Figure 2c" on the same page are lowercase, as are the panel labels in Figures 1 and 2. The targets are unambiguous, so this is cosmetic.
Second check: none (judge only)
Report a problem with this finding
Typo Duplicated word: "the effect of a protein-truncating variant (PTV) variant is eliminated by an adjacent SNP".
- p. 4 MNP paragraph
truncating variant (PTV) variant is eliminated
- p. 4 MNP paragraph
protein-truncating variant (PTV) variant is eliminated
Judge's note: Read on the page 4 image; "variant" appears before and after the abbreviation "(PTV)". Not an extraction artefact.
Second check: none (judge only)
Report a problem with this finding
Typo The LoF-intolerance threshold is printed as "pLI ≥0.9" in the definition (page 5), the Figure 3b legend and the Extended Data Figure 4 legend, but as "PTV-constrained genes (pLI >0.9)" and "constrained (pLI > 0.9) genes (Extended Data Figure 4)" on page 7.
- p. 5 Inferring variant deleteriousness, para 3
LoF intolerant (pLI ≥0.9, n=3,230)
- p. 7 Impact of rare protein-truncating variants, para 2
constrained (pLI > 0.9) genes (Extended Data Figure 4)
- p. 15 Extended Data Figure 4 legend
limited to constrained (pLI ≥0.9) genes
- p. 7 Impact of rare protein-truncating variants, para 2
PTV-constrained genes (pLI >0.9)
Judge's note: Checked the symbols on the page images, not the text layer: page 5 prints ≥ (twice, including "PTV (pLI ≥0.9)"), page 15 and page 29 legends print ≥, the Figure 3b axis prints ">= 0.9"; page 7 prints a strict > both times, and the Figure 3f key prints ">0.9". For a continuous score the gene sets are effectively identical, so this is a notation inconsistency only.
Second check: none (judge only)
Report a problem with this finding
Typo Missing article: "natural selection does not tolerate a 50% deficit in expression due to the loss of single allele" (should read "a single allele").
- p. 5 Constraint section
due to the loss of single allele
Judge's note: Read on the page 5 image; the article is absent in the printed text.
Second check: none (judge only)
Report a problem with this finding
Typo Discussion: "disease genes—particularly those that act after post-reproductive age—do not necessarily have high pLI values" is a doubled construction; the intended sense is "after reproductive age" or "at post-reproductive age".
- p. 8 Discussion, paragraph 2
particularly those that act after post-reproductive age
Judge's note: Read the crop of PDF page 8. The phrase is printed as quoted. "After post-reproductive age" literally means after the post-reproductive period; the meaning is recoverable from context, so this is cosmetic wording.
Second check: none (judge only)
Report a problem with this finding
Typo Extended Data Figure 3 title ends without its noun: "correlations between observed and expected variant counts for synonymous, missense, and protein-truncating".
- p. 14 Extended Data Figure 3 title
for synonymous, missense, and protein-truncating
Judge's note: Read the full image of PDF page 14. The bold title ends at "protein-truncating" with no noun following. The legend body uses the same ellipsis ("depicted for synonymous (b), missense (c), and protein-truncating (d)"), so this may be a deliberate elliptical style with "variants" understood from "variant counts"; either way it is cosmetic.
Second check: none (judge only)
Report a problem with this finding
Typo Affiliation 4 prints "Childrens Hospital at Westmead" without an apostrophe, while affiliation 7 prints "Boston Children's Hospital".
- p. 18 Affiliation 42
Karolinska Institute, Stockholm
- p. 19 Affiliation 56
Karolinska Institutet, Stockholm
- p. 17 Affiliation 4
Childrens Hospital at Westmead
Judge's note: Read the crop of PDF page 17: affiliation 4 prints "Childrens Hospital at Westmead" and affiliation 7 prints "Boston Children's Hospital". The apostrophe is missing in one and present in the other. They are different institutions, so the inconsistency is only suggestive; whether the Westmead hospital's own name carries an apostrophe is outside knowledge. The Karolinska Institute / Institutet part of this cluster is the same defect as C021 and is confirmed there; this entry is kept for the "Childrens" point only.
Second check: none (judge only)
Report a problem with this finding
Typo Accented characters in the affiliations are printed as substitute glyphs: "Gen—mica", "M_dicas y Nutrici—n", "Salvador Zubir‡n", "Malm_", "Technische Universit_t MŸnchen", "Aut—noma", "G_nome Qu_bec".
- p. 18 Affiliations 29-30
Instituto Nacional de Medicina Gen—mica
- p. 18 Affiliation 30
M_dicas y Nutrici—n
- p. 20 Affiliation 62
Nutrici—n Salvador Zubir‡n
- p. 21 Affiliation 89
Malm_, Sweden
Judge's note: Checked the page images, not only the text layer. PDF page 18 shows "Gen—mica" (aff. 29) and "M_dicas y Nutrici—n" (aff. 30); page 20 shows "M_dicas y Nutrici—n Salvador Zubir‡n" (aff. 62); page 21 shows "Malm_" (aff. 89) and "Universit_t MŸnchen" (aff. 103); page 22 shows "Aut—noma" (aff. 107) and "G_nome Qu_bec" (aff. 121). The corrupted glyphs are printed on the page, so this is an encoding fault in the published author manuscript and not an extraction artefact. Accented names in the collaborator list on page 19 (Jiménez, González) print correctly, which confirms the fault is specific to the affiliation text.
Second check: none (judge only)
Report a problem with this finding
Typo The same department is listed as two separate affiliations with the institution spelled two ways: affiliation 42 "Department of Medical Epidemiology and Biostatistics, Karolinska Institute, Stockholm, Sweden" and affiliation 56 "Department of Medical Epidemiology and Biostatistics, Karolinska Institutet, Stockholm, Sweden".
- p. 18 Affiliation 42
Department of Medical Epidemiology and Biostatistics, Karolinska Institute,
- p. 19 Affiliation 56
Department of Medical Epidemiology and Biostatistics, Karolinska Institutet,
Judge's note: Read the full images of PDF pages 18 and 19. Affiliation 42 (page 18) prints "Karolinska Institute" and affiliation 56 (page 19) prints "Karolinska Institutet", with identical department, city and country. The entry is duplicated and the spelling is inconsistent. Both spellings are in use for the institution, so this is a consistency point.
Second check: none (judge only)
Report a problem with this finding
Typo Missing space after a comma in affiliations 31 ("Sungkyunkwan University,Samsung Medical Center") and 39-41 ("State University of New York,Upstate Medical University").
- p. 18 Affiliation 31
Sungkyunkwan University,Samsung Medical
- p. 18 Affiliations 39-41
State University of New York,Upstate Medical
Judge's note: Read the full image of PDF page 18. Affiliation 31 prints "Sungkyunkwan University,Samsung Medical" and affiliations 39, 40 and 41 each print "State University of New York,Upstate Medical University" with no space after the comma, unlike the rest of the list. Cosmetic.
Second check: none (judge only)
Report a problem with this finding
Typo The list headed "Collaborators (alphabetical order)" is not in alphabetical order by surname: "Christopher P Jenkinson61, Jennifer Moran2, Silvia Jiménez-Morale78" places Moran among the J surnames, and "María E González-Villalpando87, Clicerio González-Villalpando87" puts María before Clicerio.
- p. 19 Collaborators list
Christopher P Jenkinson61, Jennifer Moran2, Silvia Jiménez-Morale78
- p. 19 Collaborators list
María E González-Villalpando87, Clicerio González-
- p. 19 Collaborators list
Silvia Jiménez-Morale78
- p. 18 Affiliation 42
Karolinska Institute
Judge's note: Read the full image of PDF page 19. The heading states alphabetical order and the list runs by surname (Abboud, Abecasis, ... Wilkens, Williams). "Jennifer Moran" sits between Jenkinson and Jiménez-Morale, where a first-name sort would put her, instead of among the M surnames. The two González-Villalpando entries are in reverse given-name order, a lesser point. This entry is confirmed for the ordering fault. The other two points bundled in this cluster are covered elsewhere: the "Jiménez-Morale" spelling in C027 and the Karolinska Institute / Institutet inconsistency in C021.
Second check: none (judge only)
Report a problem with this finding
Typo Collaborator listed twice in succession with different affiliation numbers: "Graeme I Bell70, Graeme I Bell70,71".
- p. 19 Collaborators list
Graeme I Bell70, Graeme I Bell70,71
- p. 19 Collaborators
Craig L Hanis91, Craig L Hanis91
Judge's note: Read the full image of PDF page 19. The collaborators list prints "Graeme I Bell" with superscript 70 and then again with superscript 70,71. An adjacent repeat of the same name and middle initial at the same affiliation is a duplicate entry. The Craig L Hanis repeat also quoted in this cluster is the subject of C025 and is confirmed there.
Second check: none (judge only)
Report a problem with this finding
Typo Collaborator listed twice in succession with the same affiliation: "Craig L Hanis91, Craig L Hanis91".
- p. 19 Collaborators list
Craig L Hanis91, Craig L Hanis91
Judge's note: Read the full image of PDF page 19. The name appears twice consecutively, both with superscript 91. An exact adjacent duplicate.
Second check: none (judge only)
Report a problem with this finding
Typo Collaborator surname printed as "Silvia Jiménez-Morale78", which looks truncated (probably "Jiménez-Morales").
- p. 19 Collaborators list
Silvia Jiménez-Morale78
Only the authors can say what is right here.
Judge's note: Rests on outside knowledge: that the collaborator's surname is Jiménez-Morales. Read the full image of PDF page 19; the name is printed "Silvia Jiménez-Morale" with superscript 78. The name appears only once in the article, so nothing in the files shows the correct spelling. "Morales" is the common Spanish surname and a dropped final letter is likely, but only the authors can confirm it.
Second check: none (judge only)
Report a problem with this finding
Typo Doubled commas in collaborator affiliation superscripts: "Martin Farrall59,,44", "Anuj Goel59,,44", "Shapour Jalilzadeh59,,44", "Theodosios Kyriakou59,,44", "Jaspal S Kooner80,,9,81", "James B Meigs102,,26", "Geoffrey Walford105,,26".
- p. 19 Collaborators list
Martin Farrall59,,44
- p. 19 Collaborators list
Jaspal S Kooner80,,9,81
- p. 19 Collaborators list
James B Meigs102,,26
- p. 19 Collaborators list
Walford105,,26
Judge's note: Read the full image of PDF page 19. The doubled commas are printed in the superscripts for Farrall, Goel, Jalilzadeh, Kyriakou, Kooner, Meigs and Walford, so they are in the published page and not introduced by text extraction. In each case the doubled comma comes where the numbering jumps from a high affiliation number back to a low one (44, 9, 26), which points to a list-merging fault. Cosmetic.
Second check: none (judge only)
Report a problem with this finding
Typo Affiliation 52 has an empty field: "Center for Non-Communicable Diseases, Karachi, , Pakistan".
- p. 19 Affiliation 52
Karachi, , Pakistan
Judge's note: Read the full image of PDF page 19. Affiliation 52 prints "Karachi, , Pakistan" with a comma, a space and a second comma between the city and the country. The stray comma is on the page.
Second check: none (judge only)
Report a problem with this finding
Typo Affiliation 108 repeats its own text and lacks spaces after commas: "Estonian Genome Centre,University of Tartu,Tartu,Estonia, University of Tartu, Tartu, Estonia".
- p. 22 Affiliation 108
Estonian Genome Centre,University of Tartu,Tartu,Estonia, University of Tartu, Tartu,
- p. 22 Affiliation 108
Estonian Genome Centre,University of Tartu,Tartu,Estonia, University of Tartu, Tartu, Estonia
Judge's note: Read the full image of PDF page 22. Affiliation 108 prints the university, city and country twice, the first time with no space after each comma. The text is malformed as printed.
Second check: none (judge only)
Report a problem with this finding
Typo Reference 18 prints the title as "Resource A Proteome-Scale Map of the Human Interactome Network"; the leading word "Resource" appears to be the journal's article-type label fused onto the title.
- p. 23 Reference 18
Resource A Proteome-Scale Map of the Human Interactome Network
- p. 23 Ref 18
Resource A Proteome-Scale Map of the Human Interactome Network.
Only the authors can say what is right here.
Judge's note: Rests on outside knowledge: that the cited Cell article's title is 'A Proteome-Scale Map of the Human Interactome Network' and that 'Resource' is Cell's article-type label. The page image (p23) shows the entry exactly as quoted: 'Rolland T, et al. Resource A Proteome-Scale Map of the Human Interactome Network. Cell. 2014; 159:1212-1226.' The wording reads as a fused section label, but the true title cannot be checked from the files.
Second check: none (judge only) · checked by hand
Report a problem with this finding
Typo Author initials are malformed in three reference entries: "Tennessen, Ja, et al." (ref 12), "Barabási, a L, Oltvai ZN" (ref 16) and "DePristo, Ma, et al." (ref 37), against the 'Surname INITIALS' form used elsewhere (e.g. "Samocha KE", "Fu W").
- p. 23 Reference 12
Tennessen, Ja, et al.
- p. 23 Reference 16
Barabási, a L, Oltvai ZN
- p. 24 Reference 37
DePristo, Ma, et al.
- p. 23 Ref 16
Barabási, a L
Judge's note: Read refs 12 and 16 on the p23 image and ref 37 on the p24 image; all three are printed as quoted, with a comma after the surname and lower-cased or split initials. Every other entry on both pages uses 'Surname AB' with no comma. The inconsistency is internal to the list and needs no outside fact.
Second check: none (judge only)
Report a problem with this finding
Typo Figure 3 legend, panel d: "are less likely to have eQTLs discovered in GTEx as the average gene" - 'as' should be 'than'.
- p. 30 Figure 3 legend, panel d
are less likely to have eQTLs discovered in GTEx as the average gene
- p. 30 Figure 3 legend, d
less likely to have eQTLs discovered in GTEx as the average gene
Judge's note: Read on the p30 image; printed as quoted. The parallel sentence for panel e on the same page uses 'more likely to be adjacent to GWAS signals than the average gene'. Grammar only; the meaning is clear.
Second check: none (judge only)
Report a problem with this finding
Typo Figure 5 legend: "the plots in c) and d) extends to 15,750 with at least one PTV and 1,550 genes with at least one homozygous PTV" - verb disagreement ('plots ... extends') and the noun 'genes' is missing after 15,750.
- p. 33 Figure 5 legend, last sentence
the plots in c) and d) extends to 15,750 with at least one PTV and 1,550 genes
- p. 33 Figure 5 legend
the plots in c) and d) extends to 15,750
Judge's note: Read on the p33 image; printed as quoted. Plural subject with singular verb, and '15,750 with at least one PTV' lacks 'genes', which the second half of the sentence supplies. Cosmetic; the meaning is recoverable.
Second check: none (judge only)