A Map of Human Genome Variation from Population-Scale Sequencing

• p. 1 Abstract

  each person carries approximately 250 to 300 loss of function variants in annotated genes

• p. 9 Results, Putative functional variants

  the 340-400 premature stops, splice site disruptions and frame shifts, affecting 250-300 genes per individual

• p. 29 Table 2, rows 'total genes containing LOF variants' (Low Coverage Individual)

  272-297

Only the authors can say what is right here.

Judge's note: Page 1 abstract and page 9 checked. Page 8 gives per-individual counts: stops 80-100, splice 40-50, frameshift 220-250. These sum to 340-400 variants. Table 2 (p29) gives 'total genes containing LOF variants' as 272-297 for a low coverage individual. The abstract's 250-300 matches the gene count while calling it a count of variants. It could be loose wording for 'variants in 250-300 genes', but read literally it gives the wrong quantity.

Second check: another model agreed

• p. 1 Abstract

  and 20,000 structural variants

• p. 28 Table 1, Union column (deletions, duplications, mobile element insertions, novel sequence insertions)

  22,025

Only the authors can say what is right here.

Judge's note: Table 1 checked on p28 (upright). Union deletions alone (22,025) exceed 20,000, and the union SV categories sum to 28,071. Page 5 gives 'over 20,000 larger SVs' for the low coverage project only (15,893 + 407 + 4,775 = 21,075). The abstract's figure seems to describe the low-coverage set rather than the whole dataset. This is an approximate figure, so I rate it minor.

Second check: another model agreed

• p. 5 Rates of variant discovery

  42x per individual across six individuals and 2.3 Gb of accessible genome

• p. 28 Table 1, Trios Total, Fraction of genome called

  2.24 Gb

• p. 4 Alignment and accessible genome

  80% of the reference, 85% of coding sequence

• p. 5 Rates of variant discovery

  2.3 Gb of accessible genome

Judge's note: Pages 4 and 5 and Table 1 on p28 checked. Similarly, the low-coverage text gives approximately 85% where the table has 86.0%. This suggests 'accessible genome' and 'fraction of genome called' may be slightly different quantities, but p4 cites Table 1 directly. The differences are small.

Second check: another model agreed

• p. 5 Rates of variant discovery

  5.9 million SNPs, 650,000 short indels

• p. 28 Table 1, Trios Total, No. of indels

  682,148

• p. 5 Rates of variant discovery

  650,000 short indels

Judge's note: Text on p5 and Table 1 on p28 checked. The text restricts its count to 1-50 bp, and the table states no size range. It may therefore be a different subset, but the paper does not say so. 682,148 would round to about 680,000.

Second check: another model agreed

• p. 11 Top paragraph

  In the latter group, only 93 (8.4%) are in strong LD

• p. 10 Last line

  nonsynonymous variants (77, 6.5%)

• p. 11 top paragraph

  only 93 (8.4%) are in strong LD

• p. 11 Application to association studies

  only 93 (8.4%) are in strong

Only the authors can say what is right here.

Judge's note: pp10-11 checked. 93/272 = 34.2% and 93/1,185 = 7.8%. 8.4% is reached only with about 1,108 (1,185 − 77 non-nonsynonymous GWAS SNPs), a base the text does not state. The authors should clarify the denominator.

Second check: another model agreed

• p. 4 Alignment and the accessible genome

  maximizing the proportion of the genome that can be interrogated

Judge's note: On the p4 image the paragraph ends with 'interrogated' and no period, so this is not an extraction loss.

Second check: none (judge only)

• p. 4 Workflow section

  publically available

Judge's note: The p4 image shows 'publically'. It is a non-standard spelling of 'publicly'. This is cosmetic.

Second check: none (judge only)

• p. 10 Application to association studies

  in strong LD with SNPs near ORDML3

• p. 10 Application to association studies

  in strong LD with SNPs near ORDML3 previously associated with asthma

• p. 10 Application to association studies

  ORDML3

Only the authors can say what is right here.

Judge's note: Rests on outside knowledge: the standard HGNC symbol is ORMDL3. The p10 image shows 'ORDML3', with the letters transposed.

Second check: none (judge only)

• p. 10 GWAS catalogue paragraph

  Of these, 1,185 (96.5%) are present

• p. 10 Application to association studies, last paragraph

  1,185 (96.5%) are present in the low coverage CEU dataset

• p. 10 Application to association studies

  1,185 (96.5%)

Judge's note: p10 checked: the denominator is 1,227 unique SNPs. The printed value is truncated rather than rounded, while the sibling percentages are rounded normally (77/1,185 = 6.5%, 272/1,185 = 23%). The error is in the last digit only.

Second check: none (judge only)

• p. 12 Positive selection section

  1000 Genome Project pilots

Judge's note: Checked the crop of p12: the page prints "1000 Genome Project pilots". The rest of the article uses "1000 Genomes Project". This is a cosmetic typo.

Second check: none (judge only)

• p. 15 Sequencing Centres list

  Washinton University in St. Louis

Judge's note: The misspelling is visible in the p15 crop. Later lists (e.g. text lines 830 and 864) spell it correctly.

Second check: none (judge only)

• p. 15 Sequencing Centres list

  Richard A. Gibbs (Principle Investigator)14

• p. 15 Broad Institute entry

  Eric S. Lander (Principal Investigator)2

Judge's note: Both spellings are visible in the p15 crops. "Principle" is the wrong word, and the list mixes it with the correct "Principal".

Second check: none (judge only)

• p. 15 Sequencing Centres, Max Planck Institute

  Wilfiried Nietfeld16

Only the authors can say what is right here.

Judge's note: Rests on outside knowledge: the author's given name is spelled "Wilfried". The p15 crop and the text layer both show "Wilfiried". The name appears only once in the files, so they cannot settle the correct spelling.

Second check: none (judge only)

• p. 16 Analysis Group, University of Washington

  Peter H. Sudmant19

• p. 17 Exon Pilot Group, University of Washington

  Sudamant19

Judge's note: Both spellings are visible in the page images, with the same affiliation 19 and the same University of Washington group.

Second check: none (judge only)

• p. 16 Analysis Group, Life Technologies

  Francisco M. De La Vega (Principle Invesigator)10

Judge's note: The misspelling is visible in the p16 crop. The text layer (line 859) shows the same misspelling in the second list.

Second check: none (judge only)

• p. 16 Analysis Group, Boston College

  Deniz Kura29l

Judge's note: Both p16 crops show "Kura29l". The page image prints it, so it is not an extraction artefact.

Second check: none (judge only)

• p. 16 Analysis Group, Wellcome Trust Sanger Institute

  Bryndis, Yngvadottir1

Judge's note: Visible in the p16 crop. The comma makes the entry read as two names.

Second check: none (judge only)

• p. 16 Sequencing Centres, Sanger

  Alison Coffey1

• p. 17 Exon Pilot Group, Sanger

  Allison Coffey1

Only the authors can say what is right here.

Judge's note: Both spellings are visible in the page crops, with the same affiliation 1. The inconsistency is internal to the paper, but which spelling is correct can only be confirmed by the authors.

Second check: none (judge only)

• p. 18 Funding acknowledgment

  ZYC200903240077A, ZYC200903240076A, and ZYC200903240080A

Only the authors can say what is right here.

Judge's note: The repeat is visible in the p18 crop. It is either a duplicate or a mistyped different grant number, and only the authors can say which.

Second check: none (judge only)

• p. 22 Figure 2 legend, panel b

  The insert shows a cartoon

Judge's note: Both wordings are visible in the p22 crop.

Second check: none (judge only)

• p. 24 Figure 3 legend

  low coverage sequencing (LowCov)

Judge's note: Checked the full p24 image: the key uses LC and the legend uses LowCov. The meaning is recoverable, so this is minor.

Second check: none (judge only)

• p. 25 Figure 4 legend b

  original Ilumina 300K

• p. 25 Figure 4b legend

  Ilumina 300K

Judge's note: Visible in the p25 crop. The main text spells it Illumina (e.g. "Illumina 1M").

Second check: none (judge only)