How to Use Single-Point Energy Calculation

Single-point energy calculation evaluates the energy of one molecular geometry without moving its atoms. It is a foundational quantum-chemistry operation used to compare conformers, rescore geometries produced by a faster method, inspect charge and spin states, and supply electronic energies to larger thermodynamic or reaction workflows.

The Neurosnap service accepts one three-dimensional SDF molecule together with its total charge and spin multiplicity. Users can choose among molecular machine-learned potentials, GFN force-field and semiempirical methods, composite 3c methods, Hartree-Fock, density-functional approximations, Skala, and DSD-BLYP-D3BJ. Materials and crystal potentials are intentionally outside this molecular service.

A single-point calculation solves or approximates the energy for the coordinates exactly as submitted. It does not optimize the geometry, search conformers, calculate frequencies, or add thermochemical corrections. Energies are therefore meaningful only when the molecular geometry, charge, multiplicity, method, and basis behavior are all specified. Absolute energies from unrelated methods should not be compared as though they share one scale.

Choose a machine-learned potential or GFN method for rapid screening, a composite 3c method for a predefined method-and-basis protocol, or a conventional Hartree-Fock or density-functional method when you need explicit basis-set control. The curated basis list covers common Basis Set Exchange choices and validates element coverage for the submitted molecule. Basis-free potentials and fixed-basis 3c methods must use the automatic setting.

Upload an SDF with the intended 3D coordinates, verify the total charge, and choose the correct multiplicity before submitting. In the result, review the resolved method and basis, convergence state, warnings, and provenance alongside the energy in hartree and electron-volts. For energy differences, calculate every structure with the same compatible protocol and preserve consistent molecular composition and electronic state unless the scientific comparison explicitly requires otherwise.

• One molecular calculation contract: The same geometry, charge, multiplicity, and result schema apply across several distinct backend families.
• Method breadth with explicit boundaries: Fast ML and GFN models, predefined composite methods, and basis-controlled electronic-structure methods are separated clearly without mixing materials models into the service.
• Curated basis selection: Common Basis Set Exchange choices feed the same validation and provenance path without requiring users to enter catalog identifiers manually.
• Interpretation-ready output: Resolved method details, convergence, warnings, and provenance appear with both hartree and electron-volt values.
• Reusable workflow stage: Results can support conformer rescoring, state comparisons, reaction studies, and later thermochemical workflows when protocols are held consistent.