The literature

This page exists so that nobody has to take the article on trust.  Every source it rests on is listed here with what it actually measured, who wrote it, and which way it cuts.

It is deliberately not a reading list of things that agree with this site.  Several mainstream sources are tagged supports, because they genuinely show that the 98.8 % slogan overstated.  Several creationist sources are tagged undercuts, because their own control rows or their own arithmetic do not reach the figure they report.

How to read an entry

Measurement or model.  A whole-genome alignment statistic is a measurement of two assemblies under one pipeline.  A percentage assembled from other people's published numbers is a recomputation, and it inherits every choice they made.  A review is neither.

Direction.  Assigned by what the source shows, not by what its author believes.

Verified.  Whether this project read the text directly (full), only its abstract or record (abstract), or only quotations of it inside another source (via).  Nothing is quoted in the main article unless it is marked full.

Field definitions

Coverage and its limits, stated plainly.  This catalogue covers the published exchange over the similarity percentage from Britten 2002 to the 2025 telomere-to-telomere comparison and its creationist and design-camp readings.  Six items could not be obtained and are listed at the bottom;  none of them blocks a check that the article depends on.

The mainstream measurements

Divergence between samples of chimpanzee and human DNA sequences is 5%, counting indels

Britten, Roy J.  PNAS 99(21):13633–13635.  2002.  https://doi.org/10.1073/pnas.172510699

camp: mainstream topic: indel-identity method: measurement direction: supports verified: full

Key figures: 779 kb in five chimpanzee BACs;  1.4 % substitution;  3.4 % in indel bases;  about 95 % shared.

The origin of the 95 % figure, three years before the consortium paper, and the only paper in the dispute that states the definitional choice as a choice:  count indel events or count indel bases.  Britten counted bases and said why.  He also stated his limits (five clones, large variation between them) and made the point the creationist literature never answers:  shared deletions are near-unforgeable markers of shared history, because a deleted stretch is not recreated by chance.  Read from the PubMed Central full text.

Initial sequence of the chimpanzee genome and comparison with the human genome

The Chimpanzee Sequencing and Analysis Consortium.  Nature 437:69–87.  2005.  https://doi.org/10.1038/nature04072

camp: mainstream topic: snv-identity, indel-identity method: measurement direction: complicates verified: full

Key figures: 1.23 % substitutions inside 2.4 Gb of best-reciprocal alignment;  1.06 % or less fixed;  about 90 Mb of indels, about 3 % of both genomes, which "dwarfs" the substitution figure;  35 million substitutions and 5 million indel events.

The source of 98.8 %, 96 % and, through its Supplementary Table S8, of Jeanson's 6 %.  All three are correct on their own definitions and all three are in the same paper.  The assembly note that "made limited use of human genome sequence (NCBI build 34) to facilitate and confirm contig linking" is what the scaffold objection rests on, and it describes ordering of contigs, not their sequence.  Read from the PDF on disk, two-column, with columns split before searching;  journal page = PDF page + 68.  Table S8 was read from the supplementary Word file, which was downloaded for the dossier.

Comparing the human and chimpanzee genomes: Searching for needles in a haystack

Varki, Ajit & Tasha K. Altheide.  Genome Research 15:1746–1758.  2005.  https://doi.org/10.1101/gr.3737405

camp: mainstream topic: indel-identity method: review direction: supports verified: full

Key figures: "not ~1%, but ~4%" in the abstract;  typical orthologous protein differs by two amino acids;  almost a third of proteins identical.

A consortium co-author put the indel-inclusive figure in an abstract three months after the consortium paper.  This is the answer to the charge that the mainstream hid the number.  It also supplies the figure the creationist side ignores in the other direction:  at the protein level the "under 1 %" holds.  Read from a Wayback Machine copy of the PDF, pages anchored from the running head.

The Evolution of Mammalian Gene Families

Demuth, Jeffery P., Tijl De Bie, Jason E. Stajich, Nello Cristianini & Matthew W. Hahn.  PLoS ONE 1(1):e85.  2006.  https://doi.org/10.1371/journal.pone.0000085

camp: mainstream topic: gene-content method: measurement direction: complicates verified: full

Key figures: 1,418 of about 22,000 human genes lack a chimpanzee ortholog, 6.4 %;  the count fell from 2,382 to 1,418 when chimpanzee coverage went from 4x to 6x;  the authors' floor is 2 %, about 400 genes.

The source of every "6 % of genes differ" claim.  It is a count of genes, not bases, and putting it beside 1.23 % is a category mistake.  The number has never been recomputed on complete genomes and its own authors expected it to fall further.  Read from PubMed Central, CC-BY.

Relative Differences: The Myth of 1%

Cohen, Jon.  Science 316(5833):1836.  2007.  https://doi.org/10.1126/science.316.5833.1836

camp: mainstream topic: epistemology method: commentary direction: supports verified: full

Key figures: relays the consortium's 1.23 % and 3 % indel figures and Demuth's 6.4 %.

A one-page News Focus piece, not research.  It is the vector by which the indel and gene-count figures entered popular argument, and the source of the Gagneux and Pääbo quotations.  Used in the article for Gagneux's sentence on there being no single measure and for its correct relay of the consortium's arithmetic.  Read from a Wayback capture of a course-site copy of the PDF, because science.org and its mirrors refuse plain fetches.

Human brain evolution: From gene discovery to phenotype discovery

Preuss, Todd M.  PNAS 109(Suppl. 1):10709–10716.  2012.  https://doi.org/10.1073/pnas.1201894109

camp: mainstream topic: indel-identity method: review direction: supports verified: full

Key figures: "not 98% or 99% identical";  "closer to 95% to 96%".

The most-quoted mainstream sentence in creationist writing on this topic.  Creationist quotations usually stop before "closer to 95% to 96%", which is the consortium's own indel-inclusive figure.  A review;  its number is cited from earlier work.  Read from PubMed Central.

High-resolution comparative analysis of great ape genomes

Kronenberg, Zev N., Ian T. Fiddes, David Gordon, et al.  Science 360(6393):eaar6343.  2018.  https://doi.org/10.1126/science.aar6343

camp: mainstream topic: structural-variation, controls method: measurement direction: complicates verified: full

Key figures: ape assemblies "humanized" by human-guided scaffolding;  chimpanzee and orangutan re-scaffolded with no human guidance;  "a modest elevation in SNV divergence";  Table 3 structural variants over 50 bp:  two humans 11.6 and 12.0 Mb, chimpanzee 82.2 Mb, gorilla 93.0 Mb, orangutan 205.1 Mb;  17,789 fixed human-specific structural variants.

Vindicates the creationist scaffold complaint as a description of the old assemblies and then tests it:  de-humanised assembly produced a modest rise in SNV divergence, not a collapse.  Table 3 is the cleanest pre-T2T human-versus-human baseline, because two long-read human genomes went through the same pipeline as the apes.  The structural-variant load is seven times the human control and orders the apes chimp, gorilla, orangutan.  Read from PubMed Central.

FluentDNA: Nucleotide Visualization of Whole Genomes, Annotations, and Alignments

Seaman, Josiah D. & Richard J. A. Buggs.  Frontiers in Genetics 11:292.  2020.  https://doi.org/10.3389/fgene.2020.00292

camp: theist-evolutionist topic: gap-inclusive, alignable-fraction method: measurement direction: undercuts verified: full

Key figures: on the UCSC hg38/panTro6 chain alignment, minus centromeres:  97.98 % aligned, 96.66 % identical;  a 200 bp "human-specific" coding segment present in chimpanzee at 98.66 % on direct BLAST;  olfactory receptor 2T10 genuinely absent from chimpanzee.

The same alignment file as Buggs 2018, from the same laboratory, twelve points higher because unassembled bases, copy-number regions and centromeres are handled differently.  The article reconciles the two by addition.  The 200 bp case is the peer-reviewed confirmation of Todd Wood's point that "unaligned" is not "different".  Carter concedes 96.6 % as the ceiling.  Read from PubMed Central, CC-BY.

Complete sequencing of ape genomes

Yoo, DongAhn, Arang Rhie, Prajna Hebbar, et al.  Nature 641:401–418.  2025.  https://doi.org/10.1038/s41586-025-08816-3

camp: mainstream topic: snv-identity, gap-inclusive, controls method: measurement direction: complicates verified: full

Key figures: 12.5–27.3 % of an ape genome fails one-to-one alignment;  Supplementary Fig. III.11 SNV divergence:  human-chimp 1.60 %, within one human 0.16 %, within one chimpanzee 0.27 %;  Supplementary Fig. III.12 gap divergence:  human-chimp 13.30 %, within one human 3.48 %, within one chimpanzee 8.09 %, within one gorilla 13.82 %;  Supplementary Table III.20 identical 1:1:  HG002 human 88.80 %, chimpanzee 83.89 %.

The most decisive source in the file, because its supplement prints the within-species controls next to the headline numbers.  Gap divergence fails its control;  SNV divergence passes it by seven times and orders the apes.  The supplement's prose on p. 25 prints the between-species SNV divergences ten times too small;  the figures two pages later are correct, and Buggs cited the figures.  The Y-chromosome rows in Tables III.19/III.20 are an aligner artefact, as Table III.17 shows.  Main paper on disk;  the 174-page Supplementary Information was downloaded for the dossier, and Figures III.11 and III.12 were rasterised and read as images because their values are rendered inside the plot.  Strengthening review, 2026-10-27:  main text p. 403 right column gives structurally divergent regions averaging 327 Mb (10 %) per ape lineage and ILS for an average of 39.5 % of the autosomes;  Supplementary Note V (SI p. 40) defines human-lineage SDRs as conserved between CHM13 and HG002 but unaligned or below 85 % identity in other apes, and Fig. V.15 (SI p. 41, raster, read via pdfimages at 2048x1116) gives 165.68 Mb human, 298.58 Mb chimpanzee, 56.23 Mb Pan ancestor, 696.18 Mb gorilla.  The human figure, about 5.3 % of the genome, corroborates the 4.91-point excess over HG002 from a different method;  both are floors.  Folio = PDF page - 1 in the SI.

Insights into hominid evolution from the gorilla genome sequence

Scally, Aylwin, Julien Y. Dutheil, LaDeana W. Hillier, et al.  Nature 483:169–175.  2012.  https://doi.org/10.1038/nature10842

camp: mainstream topic: controls, epistemology method: measurement direction: complicates verified: full

Key figures: "In 30% of the genome, gorilla is closer to human or chimpanzee than the latter are to each other" (abstract, p. 169).

Cited for the share of the genome whose local tree disagrees with the species tree, explained in the paper as incomplete lineage sorting.  It qualifies the article's ordering argument:  chimp < gorilla < orangutan is a genome-wide average, not a rule every locus obeys.  Read from the Europe PMC author manuscript on disk;  quotation confirmed against the PDF text.

Regulation and splicing

Evolution at two levels: on genes and form

Carroll, Sean B.  PLoS Biology 3(7):e245, pp. 1159–1166.  2005.  https://doi.org/10.1371/journal.pbio.0030245

camp: mainstream topic: epistemology method: review direction: supports verified: full

Key claim: King and Wilson (1975) "concluded that the small degree of molecular divergence observed could not account for the anatomical or behavioral differences between chimps and humans" (p. 1159, first column).

Cited as the route to King and Wilson's 1975 Science paper, which was not obtained.  Supports the creationist point that a sequence percentage does not measure how different the animals are;  it says nothing against common descent, which Carroll assumes.  PDF on disk, first column cropped with pdftotext -x/-W.

Global analysis of alternative splicing differences between humans and chimpanzees

Calarco, John A., Yi Xing, Mario Cáceres, et al., and Benjamin J. Blencowe.  Genes & Development 21:2963–2975.  2007.  https://doi.org/10.1101/gad.1606907

camp: mainstream topic: epistemology method: measurement direction: supports verified: full

Key figures: 6–8 % of profiled orthologous exons show pronounced (>25 percentage points) splicing differences in frontal cortex and heart (pp. 2963–2965);  about 80 % differ by under 15 points, near the platform's sensitivity limit (p. 2965);  about 4 % of profiled genes affected;  37 events checked by RT-PCR;  splicing changes fall mostly in genes whose expression level does not differ.

Supports the point that sequence identity understates regulatory difference.  Limits:  microarray, two tissues, pooled RNA, a 2007 platform.  PMC PDF via bin/pmcget.py;  folio = PDF page + 2962, anchored at PDF 2 (2964) and PDF 11 (2973).

Parallel patterns of evolution in the genomes and transcriptomes of humans and chimpanzees

Khaitovich, Philipp, Ines Hellmann, Wolfgang Enard, et al., and Svante Pääbo.  Science 309:1850–1854.  2005.  https://doi.org/10.1126/science.1108296

camp: mainstream topic: epistemology method: measurement direction: complicates verified: abstract

Key claim: "the patterns of differences in gene expression and gene sequences are markedly similar" and "consistent with a model of neutral evolution with negative selection" (abstract).

The skeptic's reply to the regulation argument:  expression divergence follows sequence divergence across five tissues, so regulation measures distance in another unit without breaking the pattern.  PubMed abstract only;  the full text is paywalled.

The theist-evolutionist critic

How similar are human and chimpanzee genomes?

Buggs, Richard J. A.  richardbuggs.com, 14 July 2018.  https://richardbuggs.com/2018/07/14/how-similar-are-human-and-chimpanzee-genomes/

camp: theist-evolutionist topic: gap-inclusive method: measurement direction: complicates verified: full

Key figures: hg38 vs panTro6:  4.06 % unaligned, 5.18 % in CNVs, 1.12 % SNPs, 0.28 % indels, about 5 % unassembled;  84.38 % one-to-one exact match.  hg38 vs panTro4:  82.34 %.

The origin of the 84 % that the creationist literature runs on, from a professor of evolutionary genomics who accepts common ancestry.  The pipeline is stated, the bins are itemised, and the post retracts his own 2008 prediction of 70 %.  A sixth of the 2018 "difference" was human DNA not yet sequenced.  Read from the page with tags stripped.

How much of a human genome is identical to a chimpanzee genome?

Buggs, Richard J. A.  richardbuggs.com, 6 May 2025.  https://richardbuggs.com/2025/05/06/how-much-of-a-human-genome-is-identical-to-a-chimpanzee-genome/

camp: theist-evolutionist topic: gap-inclusive method: recomputation direction: complicates verified: full

Key figures: 13.3 % gap plus 1.6 % SNV divergence gives 14.9 % different, 85.1 % identical.

Correct arithmetic on correctly read figures, with no within-species control applied.  The same figures give 96.4 % for one human against itself and 91.6 % for one chimpanzee against itself.  The post also misquotes its own 2018 result as 82.34 %, which was the panTro4 figure.  Read from the page with tags stripped.

Adam, Eve and Human Population Genetics, Part 10

Venema, Dennis R.  BioLogos, 2015;  original URL dead, read from the Wayback capture of 12 April 2015.

camp: theist-evolutionist topic: indel-identity, rebuttal method: commentary direction: background verified: full

Key claims: "about 95% identical, genome wide";  the 28 % of fragments excluded in a 2002 study matched more than one place, not nowhere;  Todd Wood's scholarship "entirely trustworthy".

Included because Jeanson's 11 % is built on a re-reading of this post, so the wording matters.  Venema said "genome wide";  Jeanson does not quote that word.  Buggs was right that the 95 % was a prediction rather than a measurement.  Read from the Wayback capture with tags stripped.

The creationist and design-camp sources

RTB and the Chimp Genome, part 4

Wood, Todd C.  Todd's Blog, January 2011.  https://toddcwood.blogspot.com/2011/01/rtb-and-chimp-genome-part-4.html

camp: creationist topic: alignable-fraction method: measurement direction: undercuts verified: full

Key figures: 48,975 "unalignable" chimpanzee fragments, 157 Mb, 5.2 % of the genome;  42,570 align to human by Megablast at a median 97.4 % identity;  about 25.3 Mb, 0.84 %, genuinely absent.

The decisive experiment on the central creationist premise, run by a young-earth creationist who expected the opposite result.  "Unalignable" turned out to mean "cannot be placed uniquely", not "not present".  Preliminary, unpublished, on a 6x-era assembly;  the promised full publication has not been found;  unanswered in the creationist literature.  Read from the page with tags stripped.

Genomic monkey business

Tomkins, Jeffrey P. & Jerry Bergman.  Journal of Creation 26(1):94–100.  2012.  https://creation.com/human-chimp-dna-similarity-re-evaluated

camp: creationist topic: yec-argument method: recomputation direction: complicates verified: full

Key figures: "not more than ~87% identical, and possibly not higher than 81%";  Liu et al. 98.9 % recomputed to 95.6 % with indels;  Ebersberger et al. recomputed to 65 %.

Half legitimate and half not, in the same table.  Adding indel bases to a substitution-only figure is Britten's move and gets Britten's answer.  Converting "failed to align" to "different" assumes what Wood tested and refuted.  The review dismisses Wood without engaging his result.  Read from the creation.com web version.

Comprehensive Analysis of Chimpanzee and Human Chromosomes Reveals Average DNA Similarity of 70%

Tomkins, Jeffrey P.  Answers Research Journal 6:63–69.  2013.  https://answersresearchjournal.org/chimpanzee-and-human-chromosomes/

camp: creationist topic: alignable-fraction method: measurement direction: undercuts verified: full

Key figures: 66–76 % per autosome, 70 % genome-wide, as the share of chimpanzee slices that ended inside an optimal BLASTN alignment;  parameters include -ungapped.

Two independent defects:  the ungapped aligner, which stops at the first indel and so measures distance to the next indel, and a BLASTN batch bug that Tomkins himself found and published in 2015.  The 70 % was withdrawn by its author and is still quoted in secondary creationist literature.  Read from the ARJ web version.

Documented Anomaly in Recent Versions of the BLASTN Algorithm

Tomkins, Jeffrey P.  Answers Research Journal 8:379–390.  2015.  https://answersresearchjournal.org/blastn-algorithm-anomaly/

camp: creationist topic: alignable-fraction method: measurement direction: complicates verified: full

Key figures: BLASTN versions after 2.2.25+ silently dropped queries in large batches;  reanalysis gives 88 % with BLASTN and nucmer, 73 % with LASTZ.

The most creditable paper in the creationist stack:  Tomkins found a bug that destroyed his own headline result and published it, naming the affected paper.  The reanalysis still carries ungapped, the nucmer run keeps repeat and paralogue matches with no one-to-one filter, and three aligners on the same genomes returning 73 %, 88 % and, in Buggs' hands, 98.6 % shows that the parameters produce the spread.  Read from the ARJ web version.

Analysis of 101 Chimpanzee Trace Read Data Sets

Tomkins, Jeffrey P.  Answers Research Journal 9:294–298.  2016.  https://assets.answersresearchjournal.org/doc/v9/101_chimpanzee_human_dna.pdf

camp: creationist topic: yec-argument method: measurement direction: complicates verified: full

Key figures: early trace-read batches match human about 7 % better than later ones;  concludes about 85 % identical overall.

Rests on a real phenomenon, human contamination of early sequence databases, and reports no control:  the same batches were not run against the chimpanzee genome to see whether better reads simply align better to everything.  Its statement that all trace reads "have been assembled onto the human genome" misstates the consortium paper it cites.  Contamination is not a worry for the 2025 long-read assemblies, which give a higher SNV divergence.  Read from the ARJ PDF.

Finding Adam in the Genome, Part 3

Jeanson, Nathaniel T.  Answers in Genesis, 2017.  https://answersingenesis.org/genetics/finding-adam-in-genome-response-to-chapter-2-adam-and-the-genome-part-3/

camp: creationist topic: yec-argument method: recomputation direction: undercuts verified: full

Key figures: Table S8 reports 6 % of human DNA un-aligned;  6 % plus Venema's 5 % gives 11 %, over 330 million letters.

The citation is accurate and was checked against Table S8 directly.  The inference is not:  the same table gives 20 % un-aligned for the X and 38 % for the Y, chromosomes sequenced at half the autosomal depth, so the un-aligned fraction tracks coverage.  The addition also removes the unaligned fraction from a figure Venema stated as genome-wide, the same double subtraction as Tomkins 2018.  Read from the page with tags stripped.

Comparison of 18,000 De Novo Assembled Chimpanzee Contigs to the Human Genome

Tomkins, Jeffrey P.  Answers Research Journal 11:205–209.  2018.  https://answersresearchjournal.org/comparison-chimp-contigs-human-genome/

camp: creationist topic: gap-inclusive, controls method: measurement direction: undercuts verified: full

Key figures: Table 1 mean BLASTN identity:  human 84.39 %, panTro4 91.91 %, panTro5 100.0 %;  "close to 80%" after subtracting Buggs' 4.06 %.

The source of the widely quoted 80 %, which is a double subtraction:  Buggs' 84.38 % already excludes the 4.06 %.  The paper's own Table 1 carries the control that refutes its conclusion:  Clint's contigs match an assembly of Clint's own reads at only 91.91 %, so the species-attributable difference under this pipeline is at most 7.5 points.  The panTro5 "control" was built from the very contigs queried.  Its methods set max_target_seqs 1, max_hsps 1, so one local alignment is reported per contig;  the "only about one-third of each chimpanzee contig" finding measures that one alignment, not how much of the contig can align.  Tomkins states in his Summary that the 84.4 % excludes non-alignable regions.  Read from the ARJ web version.

Human vs Chimp: an honest evaluation of our genetic differences

Carter, Robert W.  BiblicalGenetics.com, 15 August 2023.  https://biblicalgenetics.com/human-vs-chimp-differences/

camp: creationist topic: yec-argument method: commentary direction: complicates verified: full

Key claims: the highest defensible published similarity is Seaman and Buggs' 96.6 % and the true figure is below it;  the real argument is the mutation rate;  human-to-human ungapped BLAST results in the 80s are "highly suspicious".

The most honest creationist statement of the question and the one the article engages.  Carter concedes the ceiling, is the only creationist writer who has noticed the within-human control, and dismisses the decisive human-versus-human computation rather than running it.  Two citation slips in his reference list are noted in the article's footnote.  Read from the page with tags stripped.

Additional Method of Analysis Confirms Human-Chimp Genomes Are About 15 Percent Different

Luskin, Casey.  Evolution News & Science Today, May 2025.  https://evolutionnews.org/2025/05/additional-method-of-analysis-confirms-human-chimp-genomes-are-about-15-percent-different/

camp: id-adjacent topic: gap-inclusive, controls method: recomputation direction: undercuts verified: full

Key figures: 15.05 % haploid, 16.11 % diploid, 20.12 % X, 95.68 % Y, all read as "different" from the chimpanzee rows of Supplementary Tables III.19 and III.20.

Every number is transcribed correctly and every conclusion is wrong for one reason:  the human control rows are the first rows of the tables he cites.  HG002 scores 88.80 % where chimpanzee scores 83.89 %;  a human Y scores 27.02 % where chimpanzee scores 4.32 %, and 99.94 % under a different aligner.  The most-read statement of the design-camp position on the 2025 data, which is why the article answers it by name.  Read from the page with tags stripped.

The chimpanzee genome and the problem of biological similarity

Wood, Todd Charles.  Occasional Papers of the Baraminology Study Group 7:1–18.  2006.

camp: creationist topic: yec-argument, epistemology method: review direction: complicates verified: full

Key claims: common design "could be a reasonable first step" for functional genes but not for shared pseudogene lesions (p. 9);  differences are outnumbered by similarities "at least 9 to 1" (p. 9);  "Creation biology needs an explanation of the pattern of similarities, not merely an ad hoc appeal to a common designer" and ReMine's message theory is "the only recent attempt" (p. 11);  "Even at 99% identity" the source of human-chimp differences "is not likely to be found entirely in the genome sequences" (p. 12).

The primary young-earth statement that the similarity is real and that creation biology lacks an account of its pattern.  Cited for that position in Wood's own words, for the shared-deletion concession, and for the regulation point.  Wayback copy on disk;  printed page = PDF page - 2;  each quotation confirmed on its PDF page.  The p. 7 sentence "To say that God could have created the pattern is merely ad hoc" belongs to Wood's summary of the evolutionary argument, not his own claim, and is not used.

The Dependency Graph of Life

Ewert, Winston.  BIO-Complexity 2018(3):1–27.  2018.  https://doi.org/10.5048/BIO-C.2018.3

camp: id-adjacent topic: yec-argument, epistemology method: measurement direction: complicates verified: full

Key figures: the dependency graph beats a tree on all nine gene-family databases in Table 4, by over 10,000 bits on the least favourable, HomoloGene (p. 10);  the tree beats the dependency graph on the five EvolSimulator datasets (Table 3);  data are presence or absence of gene families only, with "exact sequences ... left to future research" (p. 5).

The most developed design account of nested similarity, so the article's former claim that none had been written was out of date.  It does not reach substitution divergence, which is the ordering the article reports, and Ewert says so.  CC BY, PDF on disk.

The epistemology

What It Means to Be 98% Chimpanzee

Marks, Jonathan.  University of California Press, 2002, paperback edition.

camp: mainstream topic: epistemology method: commentary direction: complicates verified: full

Key claims: two random sequences match at about 25 %, so 25 % is the zero of the scale;  humans are about 35 % daffodil;  against a snail, human and chimpanzee are over 99 % identical on every axis.

The standard mainstream argument that a percentage carries almost none of the meaning loaded onto it.  It cuts both ways:  a high figure is less impressive than it sounds and a low figure is less damaging.  Creationists quote Marks against 98 %;  his thesis is that humans and chimpanzees are obviously nearly identical and the question is what that means.  Read through archive.org's search-inside endpoint on a print-disabled item;  printed page = leaf − 24, anchored twice.  In copyright, nothing stored.

Considered and not used

• Cohen 2007, the Pääbo quotation ("I don't think there's any way to calculate a number").  Verified and on disk, not used, because it is the most abused sentence in this literature in both directions and the Gagneux sentence makes the same point without the ambiguity.

• Buggs 2025, the X and Y figures (4.18 % and 75.6 % gap divergence).  Verified, cut for length;  the article treats the sex chromosomes through Luskin's reading and Yoo's Table III.20 instead.

• Marks 2002, the paperback preface on Britten.  The closing figure in the sentence was not recovered from the OCR, so the sentence is not quoted;  Britten 2002 is cited directly.  Listed in the quarantine page.

• Yoo 2025, Supplementary Note IV orangutan comparison (99.63 % under Cactus, 97.17 % under LASTZ for the same two genomes).  Verified, not used for length;  it is the cleanest single demonstration that the aligner sets the number and belongs in a future revision.

• Eric's notes, the Marques-Bonet press quotation ("the true overall difference is likely" higher).  A press interview with no verified locator;  not used.  Listed in the quarantine page.

• Barbosa-Morais et al. 2012, Science 338:1587, "The evolutionary landscape of alternative splicing in vertebrate species."  Abstract only (PubMed, saved as barbosa-morais-2012):  "Within 6 million years, the splicing profiles of physiologically equivalent organs diverged such that they are more strongly related to the identity of a species than they are to organ type."  Not used, because the abstract does not name the species pair behind "6 million years" and the full text is paywalled, so it cannot be tied to human and chimpanzee from the bytes held.

Could not be obtained

• King, Mary-Claire and Allan C. Wilson.  "Evolution at two levels in humans and chimpanzees."  Science 188:107–116, 1975.  Science paywall;  PubMed carries no abstract;  the Memorial University PDF that search returns is a course handout, not the paper.  Cited through Carroll 2005's summary.

• Wood's promised full publication of the Megablast analysis.  Searched in Answers Research Journal, the BSG/CORE proceedings and Wood's blog index;  not found.  The 2011 figures are cited as preliminary.

• Tomkins, J. P.  "Genome-wide DNA alignment similarity (identity) for 40,000 chimpanzee DNA sequences queried against the human genome is 86–89%."  Answers Research Journal 4:233–241, 2011.  The ARJ slug patterns that work for the other papers 404.  Its headline range is quoted in the 2012 review, which is on disk.

• Glenn Williamson's analysis of Tomkins' BLASTN parameters (a Dropbox PDF).  Not fetched;  the -ungapped point is verifiable from Tomkins' own published parameter lists.

• Ebersberger et al. 2002, American Journal of Human Genetics 70:1490.  Not checked against the Tomkins and Bergman recomputation to 65 %.  PMC should hold it.

• Carter's Facebook post on the 2025 data (3.6 % within one human).  Not fetchable without a login;  the number is reproduced from Yoo's Figures III.11 and III.12.