v3.0.6

Full Changelog: v3.0.5...v3.0.6

v3.0.5

• adds periodic dft to basic calculation
• adds gxtb_wagen2026 and starling pka methods

mode parameter in submit_tautomer_search_workflow is deprecated and has no effect. Use multistage_opt_settings to control the optimization level of theory.

Full Changelog: v3.0.4...v3.0.5

v3.0.4

Adds COSMO descriptors

Full Changelog: v3.0.3...v3.0.4

v3.0.3

Adds workflow draft submission and workflow dispatch information (which includes a runtime estimate).

Full Changelog: v3.0.2...v3.0.3

v3.0.2

• Fix rowan rdkit mkdocs

v3.0.1

Full Changelog: v3.0.0...v3.0.1

Adds webhooks, presets for basic calculation settings, bug fixes.

Breaking change: tasks is now a required field in submit_basic_calculation_workflow instead of defaulting to "optimize"

v3.0.0

Full Changelog: v2.1.16...v3.0.0

New workflows

• submit_interaction_energy_decomposition_workflow — SAPT0 non-covalent interaction energy decomposition into electrostatic, exchange, dispersion, and induction components.
• submit_solvent_dependent_conformers_workflow — conformer ensemble generation and Boltzmann-weighted free energy scoring across multiple solvents (hexane, octanol, chloroform, DMSO, water by default).
• submit_relative_binding_free_energy_graph_workflow — builds a perturbation graph connecting ligands for RBFE calculations using greedy or star-map strategies.
• submit_relative_binding_free_energy_perturbation_workflow — runs FEP simulations along RBFE perturbation graph edges. Supports fast, recommended, and rigorous presets. Includes trajectory download via download_edge_trajectories() and download_all_trajectories().
• submit_hydrogen_bond_donor_acceptor_strength_workflow — combined donor + acceptor strength prediction (complements the existing submit_hydrogen_bond_basicity_workflow).

Typed result objects

All workflows now return typed result objects with named properties instead of raw dict access. The result API follows the concurrent.futures.Future pattern:

# Block until done
result = submit_multistage_optimization_workflow(mol).result()
result.energy      # float, Hartree
result.molecule    # Molecule

# Non-blocking poll
result = workflow.result(wait=False)  # returns immediately with partial data

# Stream partial results as they update
for result in workflow.stream_result():
    print(result)  # yields at each poll interval, final result last

retrieve_workflow(uuid) also returns a typed result directly. repr() on all results includes units.

Breaking changes

None. All v2 code continues to work. The new .result() pattern is recommended over the v2 .wait_for_result() + .data dict access, but both are supported.

v2.1.16

Full Changelog: v2.1.15...v2.1.16

v.2.1.15

Full Changelog: v2.1.14...v2.1.15

v.2.1.14 - Protein MD, Analogue Docking

Full Changelog: v2.1.13...v2.1.14