Release ProPrep 1.17.0 · mjgplab/proprep

Changed

PDB Filter water analysis, metric 4: burial is now computed from geometry
alone instead of an atom count normalised by an uncalibrated constant.
Each water oxygen gets its Lee-Richards accessible area against the
protein and hetero atoms (1.4 Å probe, Bondi radii; other waters never
occlude), a bulk/enclosed flag from a flood fill of probe-accessible
space, and its nearest heavy atom. Categories are physical statements:
Clash (nearest C/N/O/S/P atom under 2.2 Å, the wwPDB close-contact
criterion), Enclosed (no path to bulk solvent), Buried (0 Å² but
bulk-connected), Exposed. The count-based burial_max_expected and the
40 Å² × (1 − burial) "estimated SASA" are gone. On 3WL2 the old score
put every enclosed cavity water at 46-68%, and nothing above 80%. The
multi-radius and directional profiles (metrics 6 and 7) keep the count and
now have their own parameter entries.
The water burial table shows a Covered column (100 × (1 − SASA / the
area of a free water)) and prints every category rule with its cutoff,
in the order the rules are applied, both in the method panel and in a
legend under the table; the panel also explains that SASA is measured on
the contact sphere while enclosure is looked for one grid cell beyond it,
which is how a water can be untouchable yet bulk-connected. The unit
symbol had been U+0172 "Ų" (U with ogonek) and now reads Å².
Metrics 6 and 7 (multi-radius and directional atom counts) state the
conventions behind their labels: the "saturation" radius is the first
step with less than 10% growth, the directional "pattern" compares the
sector-count range with the mean at 0.5× and 1.5×, and the compass glyphs
are quarters of that water's own largest sector. Each label carries its
rule, and both tables gained a legend saying so; the menu names them
atom-count profiles.
The structure fixer's alternate-location picker says which viewer colour
each alternate is drawn in ("■ red in viewer") beside its occupancy, and
the viewer's representation list is labelled "Alt A (red), occ 0.60"
instead of "Altloc A", so the colours on screen can be matched to the
numbers in the prompt.
Water analysis viewer halos and the "Ordered" category. The viewer reps
were labelled Water Cat Hbond / Water Cat Ordered; they now say the
rule and its cutoff ("Waters: ≥ 3 H-bond partners (≤ 3.5 Å)"). With the
burial metric selected the halos follow the burial categories (clash,
enclosed, SASA 0, covered ≥ 90 %, covered 50–90 %) instead of the
recommendation categories, which hid a buried water behind "Highly
connected". "Ordered" was B-factor < 30 Å²; B-factors scale with
resolution and refinement, so it is now B below the median of this
structure's protein heavy atoms, and the B-factor table shows that ratio
and the median instead of 20/40/60 Å² bins.
Water burial treats a water as one 1.4 Å sphere whether it is the probe
or a crystallographic water. The crystallographic water's oxygen had been
given Bondi's 1.52 Å, the radius of an oxygen atom inside a molecule, so
the same species had two sizes and two waters "touched" at 2.92 Å where
real ones touch at 2.80 Å. Contact distance is now 2.80 Å, the isolated
water area 98.5 Å², and the enclosure reach 3.30 Å; protein, heteroatom,
and metal occluders keep Bondi radii.
Water hydrogen-bond partners are ranked by distance alone. The previous
ranking multiplied an "angle quality" (bins at 150/120/90° scored
1.0/0.8/0.6/0.3) by a distance bin and then by 1/distance again. The
angle it measured was the one at the partner atom, X–A···O, whose ideal
value depends on hybridisation (about 120–160° at a carbonyl oxygen, about
110° at an sp³ oxygen or amine), so "linear is best" rated a textbook
Ser OG contact "acceptable" and an on-axis approach "excellent"; water
partners never got an angle at all. None of the numbers had a source.
Candidates within the 3.5 Å heavy-atom cutoff are now listed closest
first and truncated to the configured maximum; the unused
hbond_angle_cutoff parameter is gone.
The water classification cascade is gone. Every analysed water used to be
given one label (Metal-coordinating, Highly connected, Interface, Buried,
Ordered, or Bulk solvent) by a first-match-wins rule chain, and the viewer
halos followed it, so a metal-bound water that was also enclosed showed
only as metal-bound. That is a single score under another name and runs
against the rest of the analysis, which reports each metric on its own.
The viewer now draws one highlight group per fact each displayed metric
establishes (metal within cutoff; 1, 2, 3, or 4 H-bond partners; B below
the protein median; the burial categories and coverage bands; interface),
each labelled with its rule and cutoff, and a water may appear in several.
Water burial counts metal ions as occluders by default. Metal ions are
their own residue class in the burial code, and the default set was
protein,hetero, so a coordinated ion neither shielded its water nor
blocked the enclosure flood fill. Default is now protein,hetero,metal.

Fixed

Coordinated waters are part of the transformer connectivity fingerprint.
The Weisfeiler-Lehman fingerprint that lets a reused transformer tell
same-name residues apart excluded waters entirely, so water renames (say
HOH to MW1 and MW2 on a di-metal site) could never auto-resolve and always
fell to the manual prompt. Every emitter now bakes two fingerprints per
role, the anhydrous one exactly as before and a hydrated one with waters
as nodes; matching tries the hydrated bijection first and falls back to
the anhydrous one, so transformers emitted before this change, and reuse
sites whose water-metal bonds are left undefined (a restraint instead),
match exactly as they did, while water roles resolve automatically when
the bonds are defined on both sides.
Structure Viewer: colouring by "Chain ID" coloured a multi-chain
structure almost uniformly red. The option used NGL's chainid scheme,
which keys on the parser's internal chain record (a new one at every TER
and every polymer-to-HETATM break), so 9YUQ's 16 chains became 112
records and the protein segments, first in the file, all sat at the red
end of the scale. The option now uses chainname, the chain letter;
saved scenes carrying chainid are mapped.
Replaying a session recorded from proprep --pdbid X or --pdbfile F
now reloads that structure first. The argument-driven load happens in
main() with no prompts, so such a session holds no loader interactions,
only metadata.pdb_id/pdb_file; replay printed that and then started
at the main menu with no structure, and the first module diverged. A
pdb_file recorded on another machine is found by basename in the
project directory; if it is absent the replay says which file to copy in.
Constant-pH/redox namelist variables (icnstph/solvph/ntcnstph, icnste/
solve/ntcnste, saltcon) reach every production simulation.mdin. Only
the live runner injected them; the batch directory writer, the
standalone writer, and the three SLURM writers staged the cpin and passed
-cpin -cpout -cprestrt but wrote the mdin untouched, so sander would
have run plain MD with a cpin on the command line. An imported .mdin is
now read, given its configured restraints and the titration namelist,
and written, instead of being copied verbatim.
The MD Manager's recommended engine assignment numbered NPT steps across
the whole queue, so with several structures only the first structure's
density equilibration was treated as early NPT (CPU) and every other
structure's went to the GPU with a fixed PME grid. NPT steps are now
numbered within each structure's own workflow.
Every generated microstate tLEaP script starts with logFile <microstate>_leap.log, so each build writes its own log instead of all
of them (and the info passes before them) appending to one leap.log
in the working directory. Parallel builds copy that log back beside the
inputs; the info passes log to scratch files that are removed with the
script. The log parser and the post-run message display read the
script's own log.
The titration-file step accepts several topologies at once (a number, a
comma list, a range, or all, which is the default). Residue and
initial-state choices are made on the first topology and reused for any
later one whose titratable residues match it exactly; a topology that
differs is prompted on its own. One titration_configs entry is kept per
topology alongside the legacy single titration_config.
The MD Manager's constant-pH/redox offer looks up each structure's own
titration files by prmtop name and lists every structure it applies to
before asking once. Previously it read the single most recent config, so
a set of five microstates was offered only the fifth's cpin.
Batch microstate generation offered the constant-E HEH heme library next
to the HCO library in the Topology Generator (and only HCR for the
reduced state). The batch path recorded no redox treatment, so the set
picker had nothing to filter on; choosing HEH loaded a library that
defines no HCO, tLEaP built the hemes as untyped, chargeless atoms, and
every microstate reported the same net charge. Batch microstates
enumerate redox states explicitly, one topology each, so the treatment is
fixed_E by construction; it is now recorded as such and the picker offers
only the fixed_E sets. The microstate info pass also refuses to continue
when tLEaP reports an unknown residue, instead of printing a charge that
omits it.
The per-water FreeSASA value that drove the "Buried" recommendation was
always 0.0: Result.atomArea() takes an atom index and was given selector
strings, and FreeSASA's PDB reader drops HETATM records so the water was
not in the structure being queried. Every water not otherwise categorised
was recommended as "Buried". The recommendation now uses the burial above.

Added

proprep-library: snapshot, reset and restore the user library
(~/.proprep/forcefield_params and ~/.proprep/transformers) without
ever deleting anything. snapshot [name] -m note copies the library to
~/.proprep/library_snapshots/<name>/ with a manifest; reset writes a
before_reset_<stamp> snapshot and then empties the library so a
practice run starts clean; restore <name> writes a before_restore_
snapshot and brings a snapshot back; list and status report. Settings,
keys, templates, profiles and web sessions are outside the library and
untouched. Restart ProPrep after a reset or restore.
The Structure Viewer can save and reload a scene. Save Scene in the
viewer's View Controls (or Save the scene shown in the open viewer in
the Structure Viewer menu) writes <name>.scene.json into the project
directory: every representation as shown (including ones added by hand
in the sidebar), the camera, the camera type, and the background, with
structure paths relative to the project so it travels with it. Load a
saved scene lists the files and restores the view; a scene whose
structure files are missing is refused with the paths named. Hand-added
representations also now survive an annotation refresh from the CLI,
which used to discard them.
The Topology Generator's cpin step now also generates cein and cpein
files, so a structure prepared for constant-redox MD can actually titrate.
It reads the topology and generates one file per titratable family:
cpinutil.py for the pH-titratable residues, ceinutil.py for the redox
heme HEH, and both files when both families are present. A constant_E
heme with titratable PRN propionates therefore yields a cpin and a cein,
which are complementary and are passed to sander together, following Amber
tutorial 33. Generating only one family is still offered, for holding the
heme at a fixed oxidation state or the propionates at fixed protonation.
cpeinutil.py is used only for a residue that titrates in proton and
electron on the same site, which neither of the other two will accept. The
MD Manager carries every generated file through to the run scripts, emitting
each one's flag triple and chaining each one's restarts independently, and
sets icnstph/solvph/ntcnstph and icnste/solve/ntcnste together in
one &cntrl when both apply. Constant-pH runs are unchanged.
The initial oxidation state of a redox residue can be set from a target
potential, the redox counterpart of setting protonation states from a target
pH. A potential above the residue's standard potential starts it oxidized,
below starts it reduced; HEH has Eo = -0.203 V. In a combined run the
propionates still follow pH against their pKa independently.
The bis-histidine c-type heme can be prepared for constant-redox MD. A new
redox_treatment choice sits alongside the existing propionate pH treatment
and offers constant_E beside the existing fixed_E. Under constant_E the
heme residue is named HEH, which is what AMBER's conste.lib calls it and
what ceinutil.py matches on, so the heme reaches the cein file with both its
ferric and ferrous charge vectors. The oxidation state is then chosen when the
cein is generated rather than built into the topology, so one library serves
both states, exactly as AMBER ships a single HEH unit. The redox-state
question is skipped when constant_E is chosen, since there is no state to
commit to at build time. The two treatments cross with the two propionate
treatments, so the cofactor now offers four sets: HEH with titratable PRN
for combined constant-pH/redox runs, HEH with static PRP/PRD for
redox-only runs, and the two existing HCO/HCR sets unchanged.
Self-contained installers. constructor/ builds ProPrep-X.Y.Z-<OS>-<arch>.sh
installers that carry the whole environment: ProPrep, AmberTools, MODELLER,
reduce, pdb2pqr and every other dependency. Attendees and new users run
one file and need no conda, no channel solve and no network during the
install. MODELLER comes with its placeholder license; users supply their own
key at run time (below). Built for Apple-silicon and Intel Macs and for
Linux x86-64 (which is also the Windows route through WSL2); see
INSTALL.md.
The MODELLER license key can be supplied at run time. ProPrep reads
$KEY_MODELLER or ~/.proprep/modeller_key before the first import modeller
and answers the modeller.config import itself, so no distributed artifact
ever contains a key and a wrong key is fixed by editing a file rather than
reinstalling. Outside a bundle this is a no-op when neither source exists,
so keys baked in by install_proprep.sh at install time keep working; a
runtime key wins over a baked one.
Web shell: sessions now belong to seats that outlive their websockets.
A dropped connection, a page reload or a laptop going to sleep no longer
kills the ProPrep process; the page reconnects with backoff and receives a
replay of the recent output, then carries on in the same session. A second
tab opened on the same seat displaces the first (one connection per seat, as
before).
Web shell hosted mode. proprep-web --seats N [--seats-dir DIR] [--public-url URL] serves N independent seats from one process, each with
its own working directory and a random link token persisted in seats.json
so links survive restarts. Opening /?seat=<token> sets a cookie, after
which the terminal, the docked structure viewer and the new Download
project button (a zip of the seat's directory) are all seat-scoped; a
missing or wrong token is refused. Hosted mode never auto-shuts down. The
default single-user launch is unchanged apart from reloads resuming instead
of restarting.
proprep-web is now also available as a standalone executable next to
proprep when the PyInstaller bundle is built.

Changed

Depositing parameters and saving a transformer ask for three names in a row
— library entry, parameter set, template — whose wording does not say which
level of the library each one sits at. The import wizard now shows entry
(a directory, one per molecule or site) against set (a key inside it, one per
derivation, several per entry) with the path each will occupy, and the
transformer creator says the template is the rename recipe rather than the
entry holding the parameters, naming the file it writes.
Web shell: the structure viewer announces itself to the shell over the
loopback bind address with a per-seat path, never through the public
address the browser used. That callback used to take the browser's Host
header and broke behind any reverse proxy, TLS terminator or tunnel; it is
now also refused from non-loopback callers.
Structure repair via MODELLER on a headless host (a plain ssh session) no
longer launches a text browser against the viewer page. webbrowser falls
back to lynx/w3m whenever TERM is set, which took over the terminal or
segfaulted on the WebGL page. Nothing is opened without a display or when
the chosen browser is text-mode (an explicit BROWSER still wins), and the
launch message prints the ssh -L line that makes the URL reachable from
the user's own machine. proprep-web gets the same guard.
Fe4S4 parameter sets: the Cys C-beta/S-gamma bond is now derived from each
solution's own Hessian instead of the ff14SB default that MCPB.py inserts
(about 1.9x too stiff and 0.05 A too short; the derived values are
bracketed by two independent published derivations). Two of the fifteen
broken-symmetry solutions (FO4, FR2) were re-derived from unconstrained
Hessians so all fifteen follow the same procedure; the averaged FO0/FR0/FD0
sets are recomputed accordingly.
Release tooling. docs/RELEASE_PROCEDURE.md documents every release
endpoint in order (source tag, conda channel, public snapshot repository and
its GitHub release, Zenodo, constructor installers). make check-version
now checks all seven version-carrying files, including
update_proprep_in_ambertools.sh (which had sat at 1.14.0 through two
releases) and constructor/construct.yaml; make snapshot exports the
public snapshot reproducibly.

Fixed

The titration step refuses igb and intdiel combinations that have no
reference energies instead of writing a file full of nulls. ceinutil.py
accepts any of igb 1/2/5/7/8 and intdiel 1/2, but the only redox residue
that exists has energies for igb 2/5/7/8 in implicit solvent, 2/5/7 in
explicit, and none at all for intdiel 2. The unsupported combinations were
accepted silently and produced meaningless output.
Residues that a titration utility would drop are now reported before the
file is written. All three utilities filter chain termini by comparing a
residue's atom count against ParmEd's definition and skip any mismatch with
no message and a zero exit status. That is the intent for a terminal
aspartate, but a cofactor is a different matter: HEH is one 87-atom
residue spanning the porphyrin and side-chain fragments of two cysteines and
two histidines, and a build one atom adrift simply stops titrating. Worse,
once any residue is dropped the -states list no longer matches the residue
count, so every hand-picked initial state is silently discarded in favour of
defaults. The count is now checked against ParmEd up front and again against
the generated file.
Explicit-solvent constant-redox runs get the radii-corrected topology they
need. ceinutil.py imports ParmEd's changeRadii and change actions and
never calls them; it has no -op flag at all, unlike cpinutil.py and
cpeinutil.py. When a cpin is generated alongside the cein, cpinutil.py
writes that topology and both files share it. For a redox-only run ProPrep
applies the same two actions directly and records the equivalent parmed
input in the generated script.
The titration step no longer asks for a tLEaP input file. It used that file
only to look up two paths, a leftover from when residue numbers had to be
mapped from the original PDB onto the tleap-renumbered topology; scanning
the topology directly removed that need. The topology is now selected
straight from the workspace.
A constant-pH bis-His c-type heme topology no longer hides the heme from
AMBER's redox tooling. The library named the residue HCO or HCR, and both
ceinutil.py and cpinutil.py select residues purely by the name recorded in
the topology, so the heme was passed over without comment and never titrated.
The parameters themselves were always AMBER's: the shipped unit is identical
to conste.lib's HEH atom for atom, in name, type and charge. Only the name
differed. Choose the constant_E treatment to emit HEH; the fixed-redox
sets keep HCO/HCR, which is correct for them because their oxidation state
really is fixed.
The bis-His c-type heme library now carries the four angle and improper terms
the titratable PRN propionate needs. PRN types both propionate oxygens as
OH and gives each up to two ghost hydrogens, an arrangement no standard
residue has, so it needs HO-OH-HO, OH-C-OH and two impropers that live in
frcmod.constph. The heme library's own header said it carried them and its
methodology said no constant-pH force field need be loaded separately, but
neither was true: building PRN against the heme library plus leaprc.conste
alone failed with three missing-parameter errors. The terms are copied
verbatim from frcmod.constph, and since none of them exist in parm10 and
none can form on an ordinary hydroxyl, adding them leaves every other residue
untouched.
A workflow step that pauses at a checkpoint was recorded as completed. The
step stops early to wait for a calculation run outside ProPrep and has
produced none of its artifacts, but the green tick satisfied the next step's
dependency and let the run walk past it — the failure then surfaced two steps
later, naming the step that consumed the missing file rather than the one
that never wrote it. A checkpoint is now in_progress: [n]ext offers it
again, and a step depending on it warns before running.
The modified-amino-acid RESP step reported "Missing ESP or AC file from
previous steps" without saying which, immediately after step 7 had announced
the AC file. It now names the artifact, the filename it expected, the
directory it searched, and the reason step 6 may have produced nothing.
The step-5 RESP selection was held only in memory, so a resumed session
re-selected from scratch — silently changing which conformers feed the fit
and demanding ESP single points the user had not run. It is now recorded in
the workspace, keyed by residue, and reused on resume.
An imported entry recorded its residue name from the .lib FILENAME rather
than the unit inside it. saveoff names an entry after the tLEaP variable it
was given, so a GDP.lib can hold unit gdp — and tLEaP matches the unit,
case-sensitively. The unit name is now read from the library, and a
divergence from the filename is reported.
Depositing a set could destroy another set's parameter files. Sets are
metadata keys, not directories, so every set in one redox/spin state shares a
folder, and the copy was an unguarded shutil.copy2: two sets whose files
shared a basename overwrote each other while both metadata leaves still
named the file. A colliding file is now written alongside (unless it is
byte-identical, where sharing it is what the other set already has), and a
rolled-back deposit restores anything it overwrote instead of deleting it.
Saving a transformer silently replaced any existing one of the same name —
a plain open(path, "w"), with sanitizing that collapses GDP ff94 and
gdp-ff94 onto the same file. The name is now checked before the
force-field questions, showing what is already there.
When Amber's parameter files could not be found, the import wizard listed
every atom type a frcmod declares as new — standard types like CT and OS
included — with nothing saying why. Widening is the safe direction, but it
now says the check could not be made and where it looked, since the usual
cause is running ProPrep outside the environment AmberTools is installed in.
The library-collision prompt asked "How would you like to proceed?" and
listed nothing to proceed with. options_map is metadata for the session
recorder — it never renders, and passing it suppresses Rich's inline choice
list — so the three options existed only in the source. They are printed now,
and the version-bump option names the set it would actually write. The
collision message also lost its [redox/spin] coordinates to Rich's markup
parser, which read the brackets as a style tag; exception text is escaped
before printing.
The interactive transformer creator could not address any residue in its own
table. BioPython reports a residue with no insertion code as a space, and
that space reached the editor's structure intact, while every command looks a
residue up with an empty insertion code; the comparison is exact, so
rename_atom A 302 O5' O5* answered No residue A/302 in the site. for a
residue printed one line above. Insertion codes are now normalized where the
structure is built and where it is queried.
New residue and atom names typed into the transformer creator were
force-uppercased, so a library whose tLEaP unit is lowercase could not be
matched by a rename — the editor could not perform the edit it exists to
perform, and the library had to be hand-edited instead. Names are now
recorded exactly as typed. Names being matched on may still be typed in any
case, and the operation records the structure's own spelling so replay
against a PDB still matches.
The PyInstaller bundle had drifted from the tree: scipy was excluded but
is a runtime dependency of pb_titrate and trajectory analysis, the
pb_titrate model compounds and the cluster profiles were not bundled, and
qmmm_prep was missing from the hidden imports. The bundle also carried the
build machine's modeller.config, license key included; it is now excluded.
ProPKA pKas for heme propionates were dropped entirely. ProPKA's .pka
file identifies ligand groups by residue name and atom name without a
residue number, so the summary parser skipped every PRN. The groups are now
recovered through ProPKA's Python API, which keeps each group's owning atom,
and merged additively (values parsed from the file win). Group type alone is
not the key: a C-terminal Lys carries a side-chain and a terminus group on
one residue and the terminus pKa was overwriting the side-chain value.
ProPKA pKas for residues numbered 1000 and above were dropped. The summary
column is fixed-width and loses the space between name and number at four
digits; on a six-protomer cytochrome filament this silently lost the whole
sixth protomer.
Several steps reported success they did not have: a checkpoint paused for an
external calculation was recorded as completed and the failure surfaced two
steps later naming the wrong step; the RESP step named two missing files
without saying which; a resumed modified-amino-acid session silently
re-selected conformers; an imported library entry took its residue name from
the file name rather than the unit inside it; depositing a set could
overwrite another set's files, and rollback then deleted them; saving a
transformer replaced an existing one of the same name without asking; and
the import wizard listed every atom type as new when Amber's parameter files
could not be found, without saying so.
Re-importing a parameter set that already exists asked how to proceed
without showing the choices (the options were passed only to the session
recorder, which also suppressed the inline [1/2/3]), and Rich swallowed
the bracketed name of the colliding set. The options are printed, the
version-bump choice names the set it would write, and exception text is
escaped.
The transformer creator's interactive editor could neither address nor
rename residues of an imported library: BioPython reports a missing
insertion code as a space, which never matched the empty string the
commands supply, and new names were force-uppercased, so the one rename
that would match a lowercase tLEaP unit was refused. Insertion codes are
normalized at the boundary and names are recorded verbatim.
MCPB RESP input: elements were inferred from the first letter of the atom
name when PDB columns 77-78 were absent, so ZN in MCPB.py's own 65-column
*_large.pdb read as element Z and aborted. Column justification,
monatomic-ion records and an allowlist of metal/halide names are tried
first. Stage 2 also refit every atom under its stronger restraint, damping
metal and ligating-atom charges toward zero; only the equivalenced atoms and
the heavy atom each hydrogen group hangs off are now freed.
Metal-site parameterization: a ligand parameterized standalone and again as
part of the site (renamed in the process) left both registrations in the
preprocessing lists, so tLEaP loaded two definitions of one ligand. Step 4
now drops the superseded halves and de-duplicates by real path.
The MoCo molybdenum ESP radius is pinned; tests that compare against
MCPB.py's MK radius table are skipped where AmberTools is absent instead of
reading as defects.
PDB Filter water analysis: choosing metric 4 (proximity-based burial)
crashed with 'PDBFilterWorker' object has no attribute 'parameters'.
The burial table read its radius, atom types, and weighting from the
filter worker instead of from the WaterAnalyzer that owns them; it is
now handed the analyzer like the metal and interface tables. Present
since the module was written, so burial analysis had never worked.

Installer SHA-256

bf85aa4ad1f888f97455fde19222b6d37a68a0dea255bca1080114ce87449368  ProPrep-1.17.0-MacOSX-arm64.sh
5e2dfe2d1b878d03a812cf400f1a1daac2fd240e8d528f78601276fc7324652c  ProPrep-1.17.0-Linux-x86_64.sh
384d4739a99810d0350151b549dbfaf940fd579b75cc0e25308ac7798069644f  ProPrep-1.17.0-MacOSX-x86_64.sh

Navigation Menu

Repository navigation

ProPrep 1.17.0

Choose a tag to compare

Sorry, something went wrong.

Sorry, something went wrong.

Uh oh!

No results found

Changed

Fixed

Added

Changed

Fixed

Installer SHA-256

Uh oh!