An Integrated Encyclopedia of DNA Elements in the Human Genome

• p. 1 Introduction, para 2

  3–8% of bases are under purifying (negative) selection 4–8 and therefore may be functional, although other analyses have suggested much higher estimates 9–11

• p. 8 The impact of selection on functional elements, para 1

  at least 3–8% of bases are under purifying (negative) selection 4–11

Judge's note: Both page images were checked. The reference list on p19 shows refs 9 (Pheasant & Mattick, 'Raising the estimate...') and 11 (Asthana) as higher-estimate sources. On p8 the phrase 'at least' could cover the higher estimates, so the p8 citation can be defended. Still, the two passages group the same references differently. This is a minor consistency point.

Second check: another model agreed

• p. 8 The impact of selection

  (GERP-based scores; 24 mammals8)

• p. 31 Figure 1 legend

  mean GERP score; 24 mammals8

• p. 19 Reference 8

  A high-resolution map of human evolutionary constraint using 29 mammals.

• p. 8 The impact of selection, para 3

  GERP-based scores; 24 mammals8

Only the authors can say what is right here.

Judge's note: The images of p8, the Figure 1 legend on p31 and reference 8 on p19 were checked. The species count in the text (24) differs from the cited study's title (29), and the paper does not say a 24-species subset was used. Only the authors can say which alignment underlies the scores.

Second check: another model agreed

• p. 17 Conclusions, para 3

  (8.5% of putative DNA:protein binding regions)

• p. 8 Summary of ENCODE-identified elements

  8.5% of bases are covered by either a TF binding site motif (4.6%) or a DHS footprint (5.7%)

Only the authors can say what is right here.

Judge's note: p8 crop: '8.5% of bases are covered by either a TF binding site motif (4.6%) or a DHS footprint (5.7%)'; that is a genome fraction, so the Conclusions misstate the denominator. The second reviewer point is not upheld: 4.6+5.7 exceeding the 8.5 union is consistent with overlap of their bases, and the claim about an 'original' 4.4% relies on outside knowledge not in the files.

Second check: another model agreed

• p. 12 Transcription factor co-associations

  Within the Tier 1 and 2 cell lines, we found 3,307 pairs of statistically co-associated factors

• p. 35 Figure 4 legend

  across the entire genome in K562 cells

Judge's note: The p12 text and p35 legend images were checked. The figure may be a K562 example from the pooled analysis, but neither the text nor the legend says so. This is an unresolved ambiguity rather than a clear error.

Second check: another model agreed

• p. 15 Rare variants…somatic variants

  reveals that there are mutational spectrum differences between DHSs and unannotated regions (0.06 Fisher’s Exact

Only the authors can say what is right here.

Judge's note: Crop of p15 confirms wording. No significance threshold is stated here; elsewhere the paper flags P < 0.05 (Figure 9C legend key). The wording states a difference without qualifying the 0.06 value; authors should clarify the threshold or soften the claim. Supplement not among files.

Second check: another model agreed

• p. 15 Rare variants section

  these are either non-synonymous SNPs or variants likely to induce loss of function by frame-shift, premature stop, or splice-site disruption

• p. 42 Figure 9 legend

  being non-synonymous and missense mutations for protein-coding regions

Only the authors can say what is right here.

Judge's note: Both crops checked. 'Missense' is a subset of non-synonymous, so the legend is redundant and omits the loss-of-function classes listed in the main text; the two definitions do not match.

Second check: another model agreed

• p. 15 Rare variants section

  We found similar numbers of potentially functional variants affecting protein-coding genes or affecting other ENCODE annotations

• p. 42 Figure 9A pies

  Common variants with Predicted Functional Effect (27,940)

Only the authors can say what is right here.

Judge's note: Viewed Figure 9A on p42: in both 'with predicted functional effect' pies, red visibly exceeds green, at roughly 2:1. The introduction (point 5) says non-coding is 'at least as large'. 'Similar numbers' understates the difference shown; exact counts are in Supplementary Tables K1/K2, which are not among the files.

Second check: another model agreed

• p. 17 Conclusions, para 2

  may be important for gene regulation than for biochemical function

• p. 17 Conclusions, para 2

  more information in the human genome may be important for gene regulation than for biochemical function

Only the authors can say what is right here.

Judge's note: Page 17 image: the sentence compares bases in direct gene regulation with protein-coding exons (1.2%); the same paragraph calls the 80% class 'marks' with functional properties, and gene regulation is itself defined biochemically in the paper. The contrast with 'biochemical function' does not match the comparison made; probably 'protein-coding function' was meant. Wording issue only.

Second check: another model agreed

• p. 1 footnote

  Present Address: Center for Bioinformatics and Computational Biology, 3115 Ag/Life Surge Bldg #296

• p. 29 Data Analysis Center author list

  Avinash D. Sahu 16

Judge's note: Page 1 shows the §§§§ footnote. On page 29 the author list has 'Avinash D. Sahu 16' with no symbol, although other present-address markers (e.g. ††††, ‡‡‡‡, O O O) do appear in the list. A grep of the text finds §§§§ only in the footnote. The footnote itself names Sahu, so the attribution is clear. This is a cosmetic defect in the marker.

Second check: none (judge only)

• p. 2 Introduction, highlight point 3

  as well hundreds of thousands of quiescent regions

Judge's note: The page 2 image (highlight point 3) shows the missing 'as'.

Second check: none (judge only)

• p. 3 Box 1

  DNaseI-seq

• p. 2 ENCODE data production

  TF-binding sites (ChIP-seq and DNase-seq)

Judge's note: The Box 1 image on p3 shows 'DNaseI-seq'. A grep finds 'DNase-seq' at several places in the text and 'DNase 1-seq' in a figure legend. All three names refer to the same assay, so this is a cosmetic inconsistency.

Second check: none (judge only)

• p. 3 Box 1, CAGE row

  formed at the initiationof transcription

Judge's note: The run-together is visible in the Box 1 image on p3.

Second check: none (judge only)

• p. 5 RNA section, first line

  and only31% of bases in sequenced transcripts

• p. 5

  only31% of bases in sequenced transcripts

Judge's note: The missing space is visible in the p5 image, first line.

Second check: none (judge only)

• p. 7 Chromosome-interacting regions, last paragraph

  These analyses portraya complex landscape

Judge's note: The run-together is visible in the p7 image.

Second check: none (judge only)

• p. 14 Insights into human genomic variation

  strong paternal bias for H3K79me2 and POL2RA

• p. 15 first paragraph

  strong allelic correlation between POL2RA and BCLAF1 binding

• p. 14 Insights into human genomic variation

  the POLR2A, H3K79me2, and H3K27me3 assays in the region of NACC2

• p. 15 Figure 8B discussion

  POL2RA and BCLAF1 binding

Judge's note: Page images p14 and p15 show 'POL2RA' twice; the same paragraph on p14 reads 'POLR2A, H3K79me2, and H3K27me3 assays'. Internal inconsistency; transposition typo.

Second check: none (judge only)

• p. 17 Conclusions, para 2

  one of the most reliable indication of an important biochemical event

Judge's note: Page 17 image confirms the singular noun after 'one of the most reliable'.

Second check: none (judge only)

• p. 19 Acknowledgments

  Research in the Pennachio lab was performed

• p. 27 Author list, LBNL group

  Len A. Pennachio

• p. 19 Acknowledgments

  Pennachio lab

• p. 19 Acknowledgements

  (PI: Pennacchio)

Only the authors can say what is right here.

Judge's note: Rests on outside knowledge: which spelling is correct (believed to be Pennacchio). The internal inconsistency is visible on the p19 and p27 images.

Second check: none (judge only)

• p. 22 Affiliation 5

  University ay Albany-SUNY

Judge's note: Visible on the p22 crop image.

Second check: none (judge only)

• p. 25 Affiliation 47

  #02-01, Genome,, Singapore

• p. 24 Present address §§

  CB10 1XL, UK..

Judge's note: Both visible on the p25 and p24 crop images; affiliation 48 has the same doubled comma.

Second check: none (judge only)

• p. 27 Affiliation 57

  School of Life Sciences, Tsinghua University, School of Life Sciences, Tsinghua University

• p. 26 Affiliation 53

  53Department Statistics

Judge's note: Both crops checked. p27 affiliation 57 has the duplicated phrase. p26 affiliation 53 reads 'Department Statistics' and 'Serra Mall,,'. These are cosmetic copyedit errors.

Second check: none (judge only)

• p. 27 Affiliation 60

  Room 353A, Bronxe, NY, USA

• p. 27 Affiliation 57

  School of Life Sciences, Tsinghua University, School of Life Sciences, Tsinghua University

• p. 27 Affiliation 60

  Bronxe, NY

Judge's note: Both visible on the p27 crop images. Rests on outside knowledge only for the spelling 'Bronx'; the duplicated Tsinghua text is internal.

Second check: none (judge only)

• p. 36 Figure 5 legend

  showing the expecting hypomethylation at TSSs

Judge's note: The crop of the p36 Figure 5 legend shows 'expecting hypomethylation' and 'genes bodies'. The same legend also has 'the two segmentations methods' (text layer line 1625). These are cosmetic typos.

Second check: none (judge only)

• p. 36 Figure 5 legend

  keybesides each heatmap

Judge's note: The crop of the p36 legend shows 'keybesides' run together in the rendered text, so this is not an extraction artefact.

Second check: none (judge only)

• p. 43 Figure 9 legend

  significantly enriched or supressed in somatic

Judge's note: The crop of the p43 Figure 9 legend shows 'supressed'. This is a misspelling.

Second check: none (judge only)

• p. 48 Table 3, CTCF row

  Also enriched for the cohesion components RAD21 and SMC3; CTCF is known to recruit the cohesion complex.

• p. 48 Table 3, CTCF row

  enriched for the cohesion components RAD21 and SMC3; CTCF is known to recruit the cohesion complex

Judge's note: The upright p48 image shows 'cohesion' twice in the CTCF row details. RAD21 and SMC3 are cohesin subunits, so the word is a typo and the intended meaning is clear from context.

Second check: none (judge only)

• p. 48 Table 3, PF row

  Regions that generally surrounding TSS segments (see below).

• p. 48 Table 3, PF row

  Regions that generally surrounding TSS segments

Judge's note: The upright p48 image shows the PF row details exactly as quoted. This is a grammatical slip.

Second check: none (judge only)

• p. 48 Table 3, CTCF row

  function in insulators assays

Judge's note: The upright p48 image shows 'function in insulators assays' in the CTCF row. The words are in the rendered table, so this is not an extraction join.

Second check: none (judge only)