v0.11.6

Full Changelog: v0.11.5...v0.11.6

Release v0.11.5

Full Changelog: v0.11.4...v0.11.5

Release v0.11.4

Full Changelog: v0.11.0...v0.11.4

synth-nmr 0.11.3

Security

• Removed compromised polyfill.io CDN script from MkDocs configuration to resolve supply-chain vulnerability.

synth-nmr 0.11.0

Added

• Argparse CLI: Replaced manual argument parsing with a robust argparse implementation. Added support for subcommands and improved error handling while maintaining interactive mode.
• Carbon Ring Currents: Expanded empirical chemical shift predictions (SPARTA+ fallback) to include ring current effects for Carbon-alpha (CA) and Carbon-beta (CB) atoms.

Changed

• Vectorized Trajectory Analysis: Refactored TrajectoryEnsemble to use biotite.structure.AtomArrayStack. This drastically reduces memory usage and enables high-speed vectorized ensemble averaging using NumPy.
• S² Performance: Fully vectorized the compute_s2_from_trajectory function, resulting in a ~100x speedup for large MD trajectories.

Fixed

• CLI Robustness: Improved handling of invalid input values in the CLI with clear error messages.
• Test Coverage: Added targeted tests for CLI validation, export commands, and trajectory edge cases, reaching 95% total coverage.

v0.10.0

Fixed

• Relaxation Calculations: Corrected scaling factors in $, $, and NOE calculations. Fixed a frequency term sign error for Heteronuclear NOEs in relaxation.py.

Added

• Scientific Benchmark Suite: New automated validation for Chemical Shift Index (CSI), J-couplings, RDC distributions, and relaxation parameters in tests/test_scientific_benchmarks.py.
• Dependency Management: Added requests to requirements-dev.txt for automated benchmark data retrieval.

Changed

• Code Quality: Resolved all Ruff and mypy errors to improve codebase maintainability and type safety.

synth-nmr 0.9.0

Added

• Security Policy: Added SECURITY.md for security policy and vulnerability reporting.
• Issue Templates: Updated issue templates for better bug tracking and feature requests.

Changed

• Modernized NMR Infrastructure: Vectorized core modules (J-coupling and relaxation rate calculations) using NumPy, replacing iterative approaches with highly performant arrays.
• CLI Refactoring: Refactored CLI interactive handlers to improve modularity and testability.
• Test Suite Expansion: Expanded test suite to 320 tests, achieving project-wide coverage.
• Ensemble Processing Optimization: Optimized trajectory and ensemble processing for handling large MD datasets efficiently.
• Code Quality:
  • Integrated Ruff for high-performance linting and formatting, replacing Black/Flake8.
  • Resolved mypy static analysis errors and enforced comprehensive PEP 484 type hint compliance.

Fixed

• NOE Processing: Fixed synthetic NOE processing bug in ensemble averaging logic.
• Backbone Angles: Resolved test error caused by a mismatch in backbone angles count (76) versus residue count (134).

synth-nmr 0.7.2

Fixed

• synth_nmr/relaxation.py — calculate_relaxation_rates(): Heteronuclear NOE
  values were identical for every residue (a flat horizontal line), regardless of
  backbone flexibility. Root cause: when the fast internal motion timescale tau_f
  defaults to 0, the order parameter S² cancels exactly in the ratio of the
  cross-relaxation rate to R₁, making NOE a function of τ_m and field strength only.
  Fix: a per-residue tau_f is now derived from S² using the heuristic
  tau_f = (1 − S²) × 500 ps + 50 ps (range 50–550 ps), consistent with
  MD-derived backbone timescales. Flexible residues (low S²) now correctly produce
  lower HetNOE values (approaching 0 and negative) while rigid secondary-structure
  elements (high S²) give HetNOE ≈ 0.5–0.8, matching experimental profiles.
  Reference: Lipari & Szabo (1982) J Am Chem Soc 104:4546.

synth-nmr 0.7.1

Added

• MD Trajectory / Ensemble NMR (synth_nmr/trajectory.py): New module for
  multi-frame NMR analysis. Features:
  • load_trajectory() — load NMR ensembles or MD trajectories from a list of
    AtomArray frames (or MDTraj Trajectory objects).
  • compute_s2_from_trajectory() — backbone S² order parameters via
    Lipari-Szabo vector autocorrelation ($S^2 = |\langle\hat{\mu}\rangle|^2$).
  • ensemble_average_shifts() — arithmetic mean across conformers (fast exchange).
  • ensemble_average_noes() — $r^{-6}$-weighted effective distances
    ($r_\text{eff} = \langle r^{-6}\rangle^{-1/6}$).
  • ensemble_average_rdcs() — arithmetic mean RDC over ensemble.
  • Full test suite in tests/test_trajectory.py (45 tests, all passing).
• SHIFTX2 integration (synth_nmr/chemical_shifts.py): predict_chemical_shifts()
  now uses SHIFTX2 (Han et al. 2011) when available in PATH, achieving ~0.44 ppm Cα
  RMSD vs ~0.9–1.1 ppm for the built-in SPARTA+ model. Automatic fallback to SPARTA+
  with a logged warning if SHIFTX2 is absent or crashes. New ShiftX2Predictor class
  available for direct use.
• New tutorial: docs/tutorials/ensemble_nmr_analysis.ipynb — 8-cell Google Colab
  notebook demonstrating S², ensemble-averaged shifts, NOE r⁻⁶ vs arithmetic mean, and
  RDC ensemble averaging using the 1D3Z 10-conformer Ubiquitin NMR ensemble.
• New science doc: docs/science/shiftx2.md — comprehensive guide to SHIFTX2
  integration (accuracy comparison table, installation instructions, API reference,
  fallback logic diagram, and automated test coverage table).

Changed

• synth_nmr_cli.py: CLI ensemble commands now correctly merge the nested
  {method: {res_id: {atom: shift}}} dict from predict_chemical_shifts() before
  passing to ensemble_average_shifts(). Added Dict and Tuple to typing imports.
  Added # type: ignore[attr-defined] for noe_dict.items() where mypy cannot
  resolve the type.
• docs/tutorials/advanced_observables.ipynb: Upgraded with improved
  visualisations (J-coupling coloured by secondary structure, NOE dual-panel figure
  showing pair count vs. r⁻⁶ signal power, RDC bar chart with sign colouring) and a
  new Section 5 — ensemble NOE effects with r_eff vs r_arith scatter and per-residue
  error bar chart.
• docs/tutorials/relaxation_analysis.ipynb: Upgraded to a 4-panel relaxation
  figure (R₁, R₂, hetNOE, S²) and a new Section 5 comparing heuristic S² against
  trajectory-derived S² from the 10-conformer ensemble, with Pearson correlation plot.
• mkdocs.yml: Added Ensemble NMR Analysis tutorial and SHIFTX2 Integration
  doc page to site navigation.
• README.md: Added PyPI version, Python versions, Ruff, Mypy, and License
  badges; added fourth Colab badge for Ensemble NMR Analysis tutorial; added
  pip install synth-nmr[trajectory] install variant; added MD Trajectory /
  Ensemble NMR to Features list.

Fixed

• synth_nmr/synth_nmr_cli.py (ruff E713): not args[j].lower() in (...) →
  args[j].lower() not in (...).
• tests/test_trajectory.py (ruff F841): Removed unused frames list
  comprehension in test_returns_dict_of_residue_ids.
• synth_nmr/trajectory.py (mypy no-redef): Renamed local variable frames to
  mdtraj_frames in MDTraj object loading branch.
• Tutorial notebooks: Fixed AttributeError in ensemble NOE cells —
  calculate_synthetic_noes() returns a list of dicts, not a nested dict; replaced
  incorrect raw.items() iteration with a for entry in raw: loop.