[us-commits] [ehb54/ultrascan3] 53f39f: Fix deoxynucleotide C2' hybridization: C4H1 -> C4H2

[emre brookes]

  Branch: refs/heads/ehb54-issue-976
  Home:   https://github.com/ehb54/ultrascan3
  Commit: 53f39f42b231277e413ace83e548826caa3ab7a4
      https://github.com/ehb54/ultrascan3/commit/53f39f42b231277e413ace83e548826caa3ab7a4
  Author: ehb54 <brookes at uthscsa.edu>
  Date:   2026-07-20 (Mon, 20 Jul 2026)

  Changed paths:
    M us_somo/etc/somo.residue.new

  Log Message:
  -----------
  Fix deoxynucleotide C2' hybridization: C4H1 -> C4H2

Deoxyribose C2' is bonded to C1', C3' and two hydrogens, so it is a CH2
(C4H2, 14.03). It was typed C4H1 (13.02), the ribose value, where C2'
carries O2' and only one hydrogen.

The deoxy blocks were created by copying the corresponding ribo block and
deleting the O2' atom line, but C2' was never retyped. Every deoxy residue
therefore came out 1.01 Da light, and since residue mass is the sum of its
atom hybrid masses, the error propagated into MW, vbar-weighted quantities
and bead masses for every DNA run.

Corrected in all 22 affected blocks (DA/DC/DG/DT, both the primed and the
starred atom-naming variants, with and without phosphate). Each deoxy
residue is now exactly 16.00 Da below its ribo counterpart, i.e. the loss
of one oxygen and nothing else; previously the gap was 17.01 Da:

  A ->DA   266.27 -> 250.27  (-16.00, was -17.01)
  C ->DC   242.24 -> 226.24  (-16.00, was -17.01)
  G ->DG   282.27 -> 266.27  (-16.00, was -17.01)
  U ->DT   243.22 -> 241.25  (-1.97 = -O +CH2, was -2.98)

Ribose C2' (A/C/G/U, and sugars such as LMT) is untouched.

Co-Authored-By: Claude Fable 5 <noreply at anthropic.com>


  Commit: 4ae31a8922b1bca3ae30d07dd6f66d2b5ac7c3bb
      https://github.com/ehb54/ultrascan3/commit/4ae31a8922b1bca3ae30d07dd6f66d2b5ac7c3bb
  Author: ehb54 <brookes at uthscsa.edu>
  Date:   2026-07-20 (Mon, 20 Jul 2026)

  Changed paths:
    M us_somo/etc/somo.atom.new
    M us_somo/etc/somo.hybrid.new
    M us_somo/etc/somo.residue.new
    M us_somo/etc/somo.saxs_atoms.new

  Log Message:
  -----------
  Add molybdenum cluster support to the SOMO parameter files

Adopts the additive part of Mattia's updated parameter files: support for
polyoxomolybdate clusters, which previously could not be typed at all.

  somo.hybrid      O2H02- (oxide O2-), MO+5, MO+6
  somo.atom        MO1-MO6 as MO+6, MO7/MO8 as MO+5; O2H02- oxide, O2H2
                   water and further O1H0 entries for the O1-O27 cluster
                   oxygens (38 rows)
  somo.saxs_atoms  Cromer-Mann form factors for MO+5 and MO+6, both the
                   9- and 11-parameter variants, matching the MO+3 rows
  somo.residue     two new residues, Molybdenum Cluster Subunit 1 (MO2)
                   and Subunit 2 (MO6)

Purely additive - no existing row is modified, so no current result can
change. Verified: every hybrid referenced by somo.residue resolves in
somo.hybrid; every (atom name, hybrid) pair used by the two new residues
exists in somo.atom; and the new electron counts are correct for Z minus
formal charge (O2- 10, MO+5 37, MO+6 36).

Mattia's copy of somo.residue predates 78465e20, so it is missing glucose;
GLC is deliberately retained here and only his new blocks were taken.

Co-Authored-By: Claude Fable 5 <noreply at anthropic.com>


  Commit: b112dfd5c5fe820c5ddbac261d394c8ea900d844
      https://github.com/ehb54/ultrascan3/commit/b112dfd5c5fe820c5ddbac261d394c8ea900d844
  Author: ehb54 <brookes at uthscsa.edu>
  Date:   2026-07-20 (Mon, 20 Jul 2026)

  Changed paths:
    M us_somo/etc/somo.atom.new

  Log Message:
  -----------
  Fix SAXS excluded volume for four hydroxyl entries in somo.atom

O10, O1A, O1B and O1N carried 9.13 A^3 as their O2H1 excluded volume, the
value belonging to a bare oxygen (O1H0/O2H0), rather than the 14.28 A^3 used
by every other hydroxyl. The four rows sit directly below O1H0/O2H0 rows for
the same atom names, so this reads as a copy-paste slip.

All 43 O2H1 rows in the file now use 14.28; before this change 39 did and
these 4 did not. Affects the excluded-volume term in SAXS curve computation
for structures containing those atom names.

>From Mattia's updated parameter files.

Co-Authored-By: Claude Fable 5 <noreply at anthropic.com>


  Commit: 1f0b373b0887e359bf2dd08e7ef193f0800b2f56
      https://github.com/ehb54/ultrascan3/commit/1f0b373b0887e359bf2dd08e7ef193f0800b2f56
  Author: ehb54 <brookes at uthscsa.edu>
  Date:   2026-07-20 (Mon, 20 Jul 2026)

  Changed paths:
    M us_somo/etc/somo.residue.new

  Log Message:
  -----------
  Fix NDP ionized-state vbar: 0.599 -> 0.589

NADPH (NDP) declared vbar 0.589 for the neutral form but 0.599 for both
ionized states. 0.599 is NAD's vbar, and the NDP header otherwise matches
NAD's line, so the two ionized values were carried over when the entry was
copied from NAD.

Every other multi-state residue in the file repeats the same vbar across all
of its protonation states; NDP was the only one of the 20 that did not. It
now uses 0.589 throughout, so vbar no longer jumps by 1.7% as a function of
pH for NADPH-containing models.

>From Mattia's updated parameter files.

Co-Authored-By: Claude Fable 5 <noreply at anthropic.com>


  Commit: af79233b29bed30e8f53969b0280ff14eef0aa04
      https://github.com/ehb54/ultrascan3/commit/af79233b29bed30e8f53969b0280ff14eef0aa04
  Author: ehb54 <brookes at uthscsa.edu>
  Date:   2026-07-20 (Mon, 20 Jul 2026)

  Changed paths:
    M us_somo/etc/somo.atom.new
    M us_somo/etc/somo.hybrid.new

  Log Message:
  -----------
  Set MO+3 radius to 0.69 (NEEDS CONFIRMATION FROM MATTIA)

Mattia's files change the Mo(III) radius from 0.83 to 0.69 A in both
somo.atom and somo.hybrid. Kept as the last commit on this branch because,
unlike everything before it, it changes results for existing structures:
it is not additive and not a demonstrable typo.

Evidence for: his three Mo radii, 0.69 (+3), 0.61 (+5) and 0.73 (+6), are
Shannon effective ionic radii for Mo(III) and Mo(V) six-coordinate and
Mo(VI) seven-coordinate. 0.83 corresponds to no standard Mo radius, and it
entered in 623ce9cc (2020) together with the rest of the multi-pKa support
files, with no stated source.

Evidence against: the metal radii in these tables are not uniformly Shannon
values (ZN+2 is 1.04 where Shannon gives 0.74, CA+2 is 1.26 where Shannon
gives 1.00), so the series may be on a different scale, and mixing a
seven-coordinate value for Mo(VI) with six-coordinate values for Mo(III)
and Mo(V) leaves the radius rising from +3 to +6 rather than falling.

Drop this commit if Mattia cannot confirm the provenance; the preceding
commits do not depend on it.

Co-Authored-By: Claude Fable 5 <noreply at anthropic.com>


Compare: https://github.com/ehb54/ultrascan3/compare/53f39f42b231%5E...af79233b29be

To unsubscribe from these emails, change your notification settings at https://github.com/ehb54/ultrascan3/settings/notifications