The Mutational Constraint Spectrum Quantified from Variation in 141,456 Humans
Yearbook of Paediatric Endocrinology, 2020 · paper · Biomedicine
Describes gnomAD, aggregating 125,748 exomes and 15,708 genomes, and identifies 443,769 high-confidence predicted loss-of-function variants. Using an improved mutation rate model, it classifies genes by tolerance to inactivation, which can improve gene discovery for diseases.
Checked: PDF from europepmc.org, fetched 4 Oct 2026. Page numbers are those of that PDF.
7 findings
Substance The text says "the rate of pLoF de novo variation in cases is 15-fold higher in genes belonging to the most constrained LOEUF decile" (Fig. 5a), but the Fig. 5a legend says such variants "are approximately 11-fold more likely to be found in cases compared to controls".
- p. 6 Constraint informs disease aetiologies, 2nd paragraph
is 15-fold higher in genes belonging to the most constrained LOEUF
- p. 7 Fig. 5 legend, panel a
11-fold more likely to be found in cases compared to controls
- p. 7 Fig. 5 legend, part a
are approximately 11-fold more likely to be found in cases compared to controls
- p. 7 Fig. 5a legend
approximately 11-fold more likely to be found in cases compared to controls
Judge's note: Both statements describe the same cohort (5,305 ID/DD cases, 2,179 controls), the same decile and the same variant class. On a zoomed render of Fig. 5a the first-decile pLoF point sits at 15 on the y axis, with the interval running from about 10 up off the top of the panel. The plot agrees with the text, so the legend's 11-fold is the wrong value. Found no statement that would make the two different quantities.
Second check: another model agreed · checked by hand
Report a problem with this finding
Reference or label A median count is given a percent sign: "we computed a median of 17.9 expected pLoF variants per gene (Fig. 2c) and found that 72.1% of genes have more than 10 pLoF variants ... an increase from 13.2% and 62.8%, respectively, in ExAC."
- p. 4 The LoF intolerance of human genes, 1st paragraph
an increase from 13.2% and 62.8%, respectively, in ExAC
Judge's note: The crop shows "13.2% and 62.8%, respectively, in ExAC". "Respectively" pairs 13.2 with the median of 17.9 expected variants per gene, which is a count, and 62.8% with 72.1% of genes. The percent sign after 13.2 is a wrong unit; it should read 13.2. Found nothing elsewhere in the article that makes 13.2% a percentage of anything.
Second check: another model agreed · checked by hand
Report a problem with this finding
Reference or label Fig. 3 legend misassigns panel letters: "Constrained genes are more likely to be lethal when heterozygously inactivated in mouse and cause cellular lethality when disrupted in human cells (c), whereas unconstrained genes are more likely to be tolerant of disruption in cellular models (d)."
- p. 5 Fig. 3 legend, c, d
lethality when disrupted in human cells (c), whereas unconstrained genes are
- p. 5 Validation of the LoF-intolerance score, 2nd paragraph
(mean = 0.962; t-test P = 10−78) (Fig. 3c)
- p. 5 Fig. 3 legend, c, d
cause cellular lethality when disrupted in human cells (c), whereas unconstrained genes are more likely to be tolerant of disruption in cellular models (d)
- p. 5 Fig. 3 legend, c, d
cause cellular lethality when disrupted in human cells (c), whereas unconstrained genes
Judge's note: On page 5, panel c plots only "Percentage of mouse het lethal knockout genes (%)". Panel d plots both "Cell essential" and "Cell non-essential" series. The body text cites Fig. 3c for the mouse result and Fig. 3d for both cell results. The legend puts human-cell lethality under (c), which contains no human cell data. It should read mouse (c) and human cells and cellular models (d).
Second check: another model agreed
Report a problem with this finding
Reference or label Reference 28 gives an impossible volume for the journal and year: "Walters, J. T. R. et al. The contribution of rare variants to risk of schizophrenia in individuals with and without intellectual disability. Nat. Genet. 511, 421 (2017)."
- p. 8 Reference 28
Nat. Genet. 511, 421 (2017)
- p. 8 Reference 9
Nat. Genet. 49, 806–810 (2017)
- p. 8 Reference list, ref. 28
individuals with and without intellectual disability. Nat. Genet. 511, 421 (2017).
Only the authors can say what is right here.
Judge's note: Rests on outside knowledge: the true volume and pages of this article (reviewers say Nat. Genet. 49, 1167-1173) and that 511, 421 belongs to a 2014 Nature paper. The zoomed reference list prints bold 511 followed by 421, so it is not an extraction artefact. The reference list itself shows the volume is wrong: ref. 9 is "Nat. Genet. 49, 806-810 (2017)", and refs 38, 44 and 45 put the journal at volume 50 in 2018 and 51 in 2019. The correct volume and page range cannot be established from the files.
Second check: another model agreed · checked by hand
Report a problem with this finding
Reference or label Reference 18 has an inconsistent page range: "Curr. Protoc. Bioinformatics 43, 11.10.1–11.19.33 (2013)."
- p. 8 Reference 18
43, 11.10.1–11.19.33
- p. 8 Reference 18
11.10.1–11.19.33
Only the authors can say what is right here.
Judge's note: Rests on outside knowledge: that the published range for this article is 11.10.1-11.10.33. The zoomed reference list prints "11.10.1–11.19.33". The range starts in unit 11.10 and ends in unit 11.19, which is implausible for a single unit-paginated article and points to a one-digit typo (19 for 10). The correct range cannot be verified from the files.
Second check: another model agreed
Report a problem with this finding
Reference or label Extended Data Fig. 10c legend reads 'Enrichment coefficient for each LOEUF decile across traits after controlling for brain expression and gene size', but the panel's y axis is τ* (not enrichment) and its third facet is titled 'Baseline + gene size + exon count'.
- p. 22 Extended Data Fig. 10c legend and y axis
Enrichment coefficient for each LOEUF decile across traits after controlling for brain expression and gene size
- p. 22 ED Fig. 10 caption, c
Enrichment coefficient for each LOEUF decile
- p. 22 ED Fig. 10c y-axis
τ*
Judge's note: Viewed the full page image for PDF page 22. Panel c has the y axis labelled τ* on a scale from about -2e-08 to 6e-08 with a dashed zero line, the same quantity and scale as panel a, which the legend calls 'the τ* coefficient'. Panel b is the one with a y axis labelled 'Enrichment' (0.6 to 1.8). The legend for c therefore names the wrong quantity. The facet titles are 'Baseline', 'Baseline + brain expression' and 'Baseline + gene size + exon count'; the legend mentions brain expression and gene size but not exon count, a smaller omission. The data are not affected; the label in the legend is wrong.
Second check: another model agreed
Report a problem with this finding
Wording The text says filtering in two gold-standard samples achieves precision "over 98.5% for indels in both exomes and genomes" and recall "more than 82% for indels for both exomes and genomes" (Extended Data Fig. 2), but the figure's threshold lines show lower values in several panels.
- p. 2 A high-quality catalogue of variation, 2nd paragraph
over 98.5% for indels in both exomes and genomes
- p. 2 A high-quality catalogue of variation, 2nd paragraph
more than 82% for indels for both
- p. 13 Extended Data Fig. 2, panels b, f, h (dashed threshold lines)
Exomes Indels
Only the authors can say what is right here.
Judge's note: Read the dashed threshold lines on a zoomed render of Extended Data Fig. 2; values are approximate. NA12878: exome indel precision about 98.0% (panel b), genome indel precision about 99.05% (panel d), recall about 90% and 85%. Synthetic diploid: exome indel precision about 84% and recall about 68% (panel f), genome indel precision about 91.6% and recall about 83% (panel h). The stated figures therefore fit NA12878 at best, and exome indel precision is slightly under 98.5% even there. The legend says panels f and h exclude 1-bp indels because they are poorly characterised in the synthetic sample, which may explain the lower values, but the text makes no such restriction and says "two samples". The underlying numbers are not in the article, so the authors need to confirm.
Second check: another model agreed
Report a problem with this finding
15 minor (typos and the like; most checked by the judge only)
Typo The same affiliation is listed twice under different numbers: "5Center for Genomic Medicine, Massachusetts General Hospital, Boston, MA, USA" (page 1) and "62Center for Genomic Medicine, Massachusetts General Hospital, Boston, MA, USA" (page 9).
- p. 1 Affiliation 5
5Center for Genomic Medicine, Massachusetts General Hospital, Boston, MA, USA
- p. 9 Affiliation 62
62Center for Genomic Medicine, Massachusetts General Hospital, Boston, MA, USA
Judge's note: Both crops show the identical affiliation string under numbers 5 and 62. The consortium list on page 9 reuses the title-page numbering (for example Daly 1,2,9; Kathiresan 5,45,85; Florez 1,7,61,62,63), so this is one continuous scheme with a duplicate entry. Cosmetic; no effect on content.
Second check: none (judge only)
Report a problem with this finding
Typo Affiliation sets differ between the title-page author list and the consortium list: "Benjamin M. Neale1,2,9" and "Mark J. Daly1,2,6,9" on page 1, but "Benjamin M. Neale1,2" and "Mark J. Daly1,2,9" on page 9.
- p. 1 Author list
Benjamin M. Neale1,2,9
- p. 9 Genome Aggregation Database Consortium list
Benjamin M. Neale1,2
- p. 1 Author list
Mark J. Daly1,2,6,9
- p. 9 Genome Aggregation Database Consortium list
Mark J. Daly1,2,9
Only the authors can say what is right here.
Judge's note: Read both lists in the crops. Page 1 prints Neale 1,2,9 and Daly 1,2,6,9; the consortium list on page 9 prints Neale 1,2 and Daly 1,2,9. James S. Ware is 1,10,11 in both lists, so the lists are meant to agree. Only the authors can say which set is right. MacArthur's 150 and 151 on page 1 are present-address notes and are not part of this.
Second check: none (judge only)
Report a problem with this finding
Typo Garbled wording in the pLoF definition: "(stop-gained), shift-reported transcriptional frame (frameshift), or alter the two essential splice-site nucleotides".
- p. 3 Identifying loss-of-function variants, para 1
shift-reported transcriptional frame (frameshift)
Judge's note: The page 3 image prints "shift-reported transcriptional frame (frameshift)". The phrase is not meaningful; the parallel verbs ("introduce", "alter") show a verb phrase such as "shift the reading frame" was intended. The meaning is recoverable from the parenthetical "(frameshift)".
Second check: none (judge only)
Report a problem with this finding
Typo Reference 29 title reads "Quantifying the impact of rare and uTheltra-rare coding variation across the phenotypic spectrum"; "uTheltra-rare" should be "ultra-rare".
- p. 8 Reference 29
rare and uTheltra-rare coding variation
- p. 8 Reference list, ref. 29
uTheltra-rare
- p. 8 Reference 52
Naturey 485
- p. 8
rare and uTheltra-rare coding variation across
Judge's note: Seen in the page crop of the reference list: "uTheltra-rare" is printed, with a stray "The" inside the word; ref. 30 directly below has "Ultra-rare". Not an extraction artefact. The "Naturey" point one reviewer attached here is handled in C017.
Second check: none (judge only)
Report a problem with this finding
Typo Reference 52 gives the journal as "Naturey 485, 242–245 (2012)"; it should read "Nature".
- p. 8 Reference 52
Naturey 485, 242–245 (2012)
- p. 8
Naturey 485, 242–245 (2012).
Judge's note: The crop shows "Naturey" in italics followed by bold 485. Ref. 51 immediately above prints "Nature 506" correctly, so the stray 'y' is a typo in the published page.
Second check: none (judge only)
Report a problem with this finding
Typo Reference 27 title begins with a lower-case letter: "Bamshad, M. J., Nickerson, D. A. & Chong, J. X. mendelian gene discovery: fast and furious with no end in sight."
- p. 8 Reference 27
mendelian gene discovery: fast and furious
Judge's note: Seen in the crop. Every other title in the list starts with a capital. The journal lower-cases 'mendelian' mid-title (ref. 5), but here it is the first word of the title; ref. 32 prints 'Mendelian disease'. Cosmetic.
Second check: none (judge only)
Report a problem with this finding
Typo Reference 48 is printed with a single author and no "et al.": "Finucane, H. K. Partitioning heritability by functional annotation using genome-wide association summary statistics. Nat. Genet. 47, 1228–1235 (2015)."
- p. 8 Reference 48
Finucane, H. K. Partitioning heritability by functional annotation
Only the authors can say what is right here.
Judge's note: Rests on outside knowledge: that the cited Nat. Genet. 2015 paper has many authors, so 'et al.' has been dropped. The page does print the entry with one author as quoted (crop viewed). The list does contain genuine single-author entries (ref. 25, Gravel, S.), so the form alone does not prove an error; the author list cannot be verified from the files.
Second check: none (judge only)
Report a problem with this finding
Typo Affiliation 47 reads "Department of Medicine, University of Mississippi Medical Center, Jackson, MI, USA", while affiliation 146 gives the same institution as "University of Mississippi Medical Center, Jackson, MS, USA".
- p. 9 Affiliation 47
University of Mississippi Medical Center, Jackson, MI, USA
- p. 10 Affiliation 146
Medical Center, Jackson, MS, USA
- p. 10 Consortium affiliation 146
University of Mississippi Medical Center, Jackson, MS, USA
Judge's note: Both crops viewed: affiliation 47 on PDF page 9 prints 'MI', affiliation 146 on PDF page 10 prints 'MS'. The same institution and city carry two state abbreviations, an internal inconsistency; MS is the form consistent with the institution's name (Mississippi).
Second check: none (judge only)
Report a problem with this finding
Typo Affiliation 54 reads "Centro de Investigación Biomédica en Red Enfermedades Cardiovaculares (CIBERCV)"; "Cardiovaculares" should be "Cardiovasculares".
- p. 9 Affiliation 54
Enfermedades Cardiovaculares (CIBERCV)
Judge's note: Seen in the crop of PDF page 9: 'Cardiovaculares' is printed without the 's'. Plain misspelling.
Second check: none (judge only)
Report a problem with this finding
Typo Affiliation 129 reads "Institute of Clinical Medicine, Neurology, University of Eastern Finlad, Kuopio, Finland"; "Finlad" should be "Finland".
- p. 10 Affiliation 129
University of Eastern Finlad, Kuopio, Finland
- p. 10 Affiliation 129
University of Eastern Finlad
- p. 10 Affiliation 129
University of Eastern Finlad, Kuopio
Judge's note: Seen in the crop of PDF page 10. Affiliation 79 on PDF page 9 prints 'University of Eastern Finland' correctly.
Second check: none (judge only)
Report a problem with this finding
Typo Affiliation 147 reads "Program in Infectious Disease and Mi--crobiome, Broad Institute of MIT and Harvard"; "Mi--crobiome" should be "Microbiome".
- p. 10 Affiliation 147
Program in Infectious Disease and Mi--crobiome
- p. 10 Affiliation 147
Mi--crobiome
- p. 10 Affiliation 147
Infectious Disease and Mi--crobiome
Judge's note: Seen in the crop of PDF page 10: the word sits mid-line with two hyphens inside it, so it is not a line-break hyphen or an extraction artefact.
Second check: none (judge only)
Report a problem with this finding
Typo Extended Data Fig. 3 legend: "the point at 10 shows the performance of the 10% best-scored data; the point at 50 shows the performance 50% best-scored data"; "of the" is missing in the second clause.
- p. 14 Extended Data Fig. 3 legend, a–j
the point at 50 shows the performance 50% best-scored data
Judge's note: Seen in the crop of PDF page 14. The parallel clause just before has 'performance of the 10%'; the second drops 'of the'. Missing words.
Second check: none (judge only) · checked by hand
Report a problem with this finding
Typo Extended Data Fig. 7 legend, panel h: "Centre line denotes the median; box limits denote upper and lower quartiles; whiskers denote 1.5× the interquartile range; points denote outliers)." has a closing parenthesis with no opening one.
- p. 19 Extended Data Fig. 7 legend, panel h
points denote outliers).
Judge's note: Seen in the crop of PDF page 19: the sentence before ends 'for each LOEUF decile.' with a full stop and no opening parenthesis before 'Centre line', so the ')' after 'outliers' is unmatched.
Second check: none (judge only)
Report a problem with this finding
Typo Extended Data Fig. 8a legend key reads 'weaker drug activites'; it should read 'weaker drug activities'.
- p. 20 Extended Data Fig. 8a, legend key
weaker drug activites
Judge's note: Viewed the full page image for PDF page 20. The key of panel a lists 'approved drugs', 'ChEMBL drug activities', 'weaker drug activites', 'no known ligands'. The third entry is misspelled, and the entry above it has the correct spelling. The text is inside the figure, so it is not an extraction artefact.
Second check: none (judge only)
Report a problem with this finding
Typo Reporting Summary reads 'known first disease relatives of those with severe pediatric disease' (Data exclusions, and again under Population characteristics); it should read 'first degree relatives'.
- p. 24 Reporting Summary, Data exclusions
known first disease relatives of those with severe pediatric
- p. 24 Reporting Summary, Population characteristics
known first disease relatives of those with severe
Judge's note: Viewed the crop of the Data exclusions box on PDF page 24, which prints 'known first disease relatives of those with severe pediatric disease were excluded'. The article's own methods text says 'first-degree relatives'. The second occurrence, under Population characteristics, I confirmed in the text layer only (the same phrase 'known first disease relatives of those with severe'); both crops supplied show the Data exclusions box, so I did not see the second one in an image. Wrong word, with an evident fix.
Second check: none (judge only)