CLUSTAL W: Improving the Sensitivity of Progressive Multiple Sequence Alignment through Sequence Weighting, Position-Specific Gap Penalties and Weight Matrix Choice
Nucleic Acids Research, 1994 · paper · Biomedicine
Presents CLUSTAL W, an improved progressive multiple sequence alignment program. It weights sequences, varies substitution matrices by divergence, and uses position-specific gap penalties, improving sensitivity for divergent protein sequences.
Checked: PDF from europepmc.org, fetched 5 Oct 2026. Page numbers are those of that PDF.
2 findings
Reference or label Reference 22 "Bashford, D., Chothia, C. and Lesk, A.M. (1987) J. Mol. Biol. 196, 199-216." is the same entry as reference 25, yet 22 is cited for the "fast approximate method (22)" of pairwise alignment and 25 for the "7 globin sequences of known tertiary structure (25)".
- p. 8 Reference list, entry 22
22. Bashford, D., Chothia, C. and Lesk, A.M. (1987) J. Mol. Biol. 196, 199-216.
- p. 8 Reference list, entry 25
25. Bashford, D., Chothia, C. and Lesk, A.M. (1987). J. Mol. Biol. 196, 199-216.
- p. 2 Material and Methods, The distance matrix/pairwise alignments
calculated using a fast approximate method (22)
- p. 2 Introduction, right column
alignments (22) or full dynamic programming for the distance
Only the authors can say what is right here.
Judge's note: Read the reference list on page 8: entries 22 and 25 are identical apart from a full stop after the year. Read both citations of (22) on page 2 (fast approximate alignments / fast approximate method in the original CLUSTAL programs) and the citation of (25) for the globin structures. One paper listed under two numbers is a defect in itself, and the two uses are for unrelated things, so entry 22 appears to have been replaced by a copy of entry 25. Which work was meant for 22 cannot be determined from the files (a reviewer's suggestion of Wilbur and Lipman rests on outside knowledge and is not verified here).
Second check: another model agreed
Report a problem with this finding
Reference or label Figure 4 legend: "Secondary structure assignments for the solved Spectrin (24) and Fyn (39) domains are according to DSSP (40)." No secondary structure assignments appear in the printed figure.
- p. 7 Figure 4 legend
Secondary structure assignments for the solved Spectrin (24) and Fyn (39) domains are according to DSSP (40).
Only the authors can say what is right here.
Judge's note: Looked at Figure 4 on the page 7 image: it begins with the ASV_vSRC row directly under the running head and ends with H_Atk; there is no row or annotation of secondary structure above, below or within the alignment. The legend describes an element the figure as printed does not contain; it may have been dropped in production.
Second check: another model agreed
Report a problem with this finding
11 minor (typos and the like; most checked by the judge only)
Reference or label Reference 11 is printed as "Vingron, M. and Waterman, M.S. (1993) J. Mol. Biol. 234, 1-12."; the cited paper is believed to have appeared in J. Mol. Biol. volume 235 (1994), pages 1-12.
- p. 8 Reference list, entry 11
11. Vingron, M. and Waterman, M.S. (1993) J. Mol. Biol. 234, 1-12.
Only the authors can say what is right here.
Judge's note: Rests on outside knowledge: that Vingron and Waterman, "Sequence alignment and penalty choice", was published in J. Mol. Biol. 235, 1-12 in 1994, not volume 234 in 1993. The crop of PDF page 8 confirms the entry prints "(1993) J. Mol. Biol. 234, 1-12" as quoted. Nothing in the files gives the true volume or year, so this cannot be verified here; it should be checked against Crossref or the journal before being relied on. The authors may have cited an in-press copy with a provisional volume.
Second check: another model agreed
Report a problem with this finding
Wording Introduction: improvements come "without sacrificing any of the speed and efficiency"; closing discussion: "This is achieved with almost no sacrifice in speed and efficiency."
- p. 1 Introduction, first paragraph
without sacrificing
- p. 8 Results and Discussion, final section, first lines of page
is achieved with almost no sacrifice in speed and efficiency
- p. 5 Software and algorithms, Dynamic programming
This sacrifices some
- p. 1 Introduction
any of the speed and efficiency
Judge's note: Read both sentences on the page 1 crop and the page 8 image. The absolute wording in the introduction is softened to 'almost no sacrifice' at the end; the paper gives no timings. The Myers and Miller sentence on page 5 ('This sacrifices some processing time') is about trading time for memory in the dynamic programming step and is not clearly a comparison with earlier CLUSTAL versions, so it is not counted as a contradiction. A wording inconsistency only.
Second check: another model agreed
Report a problem with this finding
Wording SH3 test set described as including "sequence pairs with as little as 12% identity" (Introduction) and as having "a minimum similarity below 12% identity" (Results).
- p. 2 Introduction, last paragraph
sequence pairs with as little as 12% identity
- p. 6 Results and Discussion, Alignment of SH3 domains
a minimum similarity below 12% identity
- p. 2 Introduction, last paragraph
as little as 12% identity
Judge's note: Both phrases read on the page images / text. The same minimum pairwise identity is put at 12% in one place and below 12% in the other; a value such as 11.x% would fit both loosely. No exact figure is given anywhere in the paper. Minor wording inconsistency; either form could be legitimate.
Second check: another model agreed
Report a problem with this finding
Wording Figure 1: the distance matrix gives Hba_Human/Hba_Horse as ".13", while the rooted tree gives their terminal branches as ".055" and ".065", which sum to 0.120.
- p. 3 Figure 1 distance matrix, row 4 col 3
.13
- p. 3 Figure 1 rooted tree, Hba branches
.055 / .065
Only the authors can say what is right here.
Judge's note: Rests on outside knowledge: in a Neighbour-Joining tree the terminal branches of two sister leaves sum to their pairwise distance. Enlarged the matrix and the rooted tree from the page 3 image: row 4, column 3 reads .13; the Hba branches read .055 and .065. Their sum, 0.120, is outside what rounds to .13 (0.125 to 0.135). The Hbb pair agrees (0.081 + 0.084 = 0.165, matrix .17). The paper does not say whether the tree and the matrix in the figure were computed from the same distances, so only the authors can say which is right; the effect is small.
Second check: another model agreed
Report a problem with this finding
Wording Text: "A single misaligned sequence, H_P55, remains in block 2 (boxed in Figure 4), while block 3 is now completely aligned"; Figure 4 legend: "The single misaligned segment in H_P55 and the misaligned residue in H_NCK/2 are boxed."
- p. 6 Results and Discussion, Alignment of SH3 domains
A single misaligned sequence, H_P55, remains in block 2 (boxed in Figure 4)
- p. 7 Figure 4 legend
The single misaligned segment in H_P55 and the misaligned residue in H_NCK/2 are boxed
- p. 6 Results and Discussion, Alignment of SH3 domains, second paragraph
A single misaligned sequence,
- p. 6 Results and Discussion, Alignment of SH3 domains, second paragraph
3 is now completely aligned
Judge's note: Read the results paragraph (crop) and Figure 4 with its legend on page 7; the figure shows a boxed residue in the H_NCK/2 row as well as the boxed segment in H_P55. The text mentions only H_P55 as still misaligned and does not mention the H_NCK/2 residue. The text can be read as counting misaligned segments within block 2 only, so this is an omission in the text, not a clear contradiction.
Second check: another model agreed
Report a problem with this finding
Typo "We increase the gap opening penalty within eight residues of exising gaps." ('exising' for 'existing')
- p. 2 Introduction, right column, first paragraph
within eight residues of exising gaps
- p. 2
residues of exising gaps
Judge's note: The page 2 image, right column, prints 'exising gaps'. Not an extraction artefact.
Second check: none (judge only)
Report a problem with this finding
Typo "Traditionally, one chooses one weight matrix and two gap penalties (one for opening a new gap and one for extending an existing gap) and hope that these will work well" ('hope' for 'hopes')
- p. 2 Introduction, left column, alignment parameter choice paragraph
one for extending an existing gap) and hope that these will work
- p. 2 Introduction, parameter choice paragraph
one chooses one weight matrix and two gap penalties (one for opening a new gap and one for extending an existing gap) and hope that these will work
Judge's note: Read on the page 2 image, left column. The subject is 'one' with 'chooses'; the coordinated verb is printed 'hope'.
Second check: none (judge only)
Report a problem with this finding
Typo Worked example of the Hbb_Human weight prints the last term as "plus one sixth the branch leading to all the vertebrate globins (0.062). This sums to a total of 0.221", where the three preceding terms show the divisor (0.226/2, 0.061/4, 0.015/5); the term should read (0.062/6).
- p. 3 The guide tree, right column
This sums to a total of 0.221
- p. 3 Figure 1, rooted tree
Hbb_Human: 0.221
- p. 3 The guide tree, worked example for Hbb_Human
globins (0.062). This sums to a total of 0.221
- p. 3 The guide tree, worked example of Hbb_Human weight
plus one sixth the branch leading to all the vertebrate globins (0.062). This sums to a total of 0.221.
Judge's note: Read the paragraph and an enlarged crop of the Figure 1 rooted tree (branches .081, .226, .061, .015, .062; weight 0.221). The parenthesis for the last term gives the whole branch length, not the divided share, unlike the other terms; with 0.062 undivided the terms give 0.274. With 0.062/6 the sum is 0.2226 against the printed 0.221. That 0.0016 gap is not confirmed as an error: the other weights in the figure differ from sums of the three-decimal branch lengths by similar amounts (Hba_Human 0.1931 vs 0.194, Glb5_Petma 0.3993 vs 0.398), which fits weights computed from unrounded lengths. Only the missing '/6' is confirmed.
Second check: none (judge only)
Report a problem with this finding
Typo Figure 3 legend: "The inital gap opening penalty is indicated by a dotted line." ('inital' for 'initial')
- p. 4 Figure 3 legend
The inital gap opening penalty is indicated by a dotted line
- p. 4 Figure 3 legend
The inital gap opening penalty
- p. 4 Figure 3 legend
The inital gap opening penalty is indicated by a dotted line.
Judge's note: The crop of the Figure 3 legend prints 'inital'. Not an extraction artefact.
Second check: none (judge only)
Report a problem with this finding
Typo Reference 36: "J. Mol. Biol. 235-1501-1531." (dash instead of a comma between volume and pages; should read 235, 1501-1531)
- p. 8 Reference list, entry 36
J. Mol. Biol. 235-1501-1531.
- p. 8 Reference list, entry 36
J. Mol. Biol. 235-1501-1531
Judge's note: The page 8 image prints '235-1501-1531' in entry 36; every other journal entry separates volume and pages with a comma (for example entry 12, 'J. Mol. Biol. 224, 461-471').
Second check: none (judge only)
Report a problem with this finding
Typo Reference 10 prints the last author as "Wooton, J.C." ("and Wooton, J.C. (1993) Science 262, 208-214."); the author of the cited Gibbs-sampling paper is believed to be J.C. Wootton, with a double t.
- p. 8 Reference list, entry 10
and Wooton, J.C. (1993) Science 262, 208-214.
Only the authors can say what is right here.
Judge's note: Rests on outside knowledge: that the last author of Lawrence et al. (1993) Science 262, 208-214 spells the surname "Wootton". The crop of the reference list on PDF page 8 clearly prints "Wooton" with a single t, so this is not an extraction artefact. The name appears nowhere else in the article, so the spelling cannot be checked from the files; it should be confirmed against the cited paper.
Second check: none (judge only)