Recipes by topic (from building a molecule to a calculation)

These recipes take one molecule each from an empty window (or a structure file, or an earlier result) to an NWChem calculation and its results, using MolForma’s interface only. Pick the topic closest to the molecule you want to build; along the way you try the features it needs (atoms and bonds, skeletons and substituents, coordination centers and ligands, peptides from a sequence, stereochemistry, point charges, BSSE correction, transition-state searches, reusing results, isotopes, your own parts and more).

The calculations on these pages were run on 2026-10-26 on a PC with an AMD Ryzen 7 2700X (8 cores, 16 threads) and 32 GB of memory, using NWChem 7.3.1 (Linux, 2–8 MPI processes). The run times on each page were measured on this PC too; they depend strongly on the CPU and its core count, the number of MPI processes and other calculations running at the same time. Another NWChem version or process count may also change the last digits of energies or the number of optimization steps.

Using these recipes

By feature“Features” describes each part of the window and what every button does, one page per screen.
By topicthe 16 recipes on this page. Each starts from a new molecule (or a structure file or a calculation result) and goes through the calculation settings, the run and the results.
The pages“What you build” → steps → calculation settings → run and results → “Troubleshooting”. Names in quotes (“Clean up all”) are as shown in the window.
Practice dataThe finished projects (.molforma.json) are linked under “Practice data” in the table of each recipe (docs/recipes/samples/ of the repository). The web manual downloads them from the links; in the app’s “Guide and manual”, clicking a link saves a copy where you choose and opens it (a part file .molforma-part.json goes to “My parts”). Use them to start halfway or to compare with your own. They carry no run settings (NWChem executable, job folder and so on), so check “3 Run settings” before running.

Built structures and calculated ones

A structure right after placing parts, or after “Clean up all”, is a starting structure for a calculation. “Clean up all” shapes the molecule with an Open Babel force field (MMFF94, UFF); it is not NWChem’s geometry optimization. To use an optimized structure for the next calculation, bring it in with “Apply as candidate” in the “Result viewer” (recipe 14).

The total charge and multiplicity of the calculation are set in “Charge and spin” of “2 Calculation”. The formal charges of atoms (the “Attributes” page) prepare the structure, for instance for adding hydrogens, and are kept apart from the total charge. When residues, fragments, coordination centers or ligands are added, a total charge that matched the value the structure suggests (the sum of formal charges, with the oxidation numbers of coordination centers) follows the new value. The multiplicity is never guessed: set it yourself for radicals and open-shell metals.

Recipes

No.	Topic	Molecule	Main features	Calculation
1	Building from atoms and bonds	Ethanol (9 atoms), acetaldehyde, dimethyl ether	Placing atoms and replacing hydrogens, deletion, bond types, inserting an atom into a bond, cleanup, measuring distances	B3LYP/6-31G* optimization → frequencies
2	Coordination compounds	trans-[Co(en)₂Cl₂]⁺ (27 atoms, charge +1)	Coordination centers, geometry and oxidation number, bidentate ligands, placing ligands one after another, the charge following them	B3LYP/def2-SVP optimization
3	Substituted aromatics	p-Nitroaniline (16 atoms)	Skeleton templates, substituents replacing hydrogens, ortho/meta/para	B3LYP/6-31G* optimization → frequencies
4	Heterocycles from SMILES and structure files	Caffeine (24 atoms)	From SMILES, 2D structure files (.sdf), adding hydrogens, Regenerate 3D	B3LYP/6-31G* optimization
5	Chains and conformers	Anti and gauche n-butane	Measuring and setting dihedrals, Save as, relaxed scans	B3LYP/6-31G* relaxed scan (13 points)
6	Stereochemistry	R- and S-lactic acid, E- and Z-2-butene	Record stereo (R/S, E/Z), Invert stereocenter, Switch E/Z	B3LYP/6-31G* optimization → frequencies (E/Z)
7	Peptides	ACE–Ala–His–Glu–NME (54 atoms, charge −1)	Build from sequence, terminal caps, side-chain states, the sequence bar	B3LYP/6-31G* single point (COSMO, water)
8	Mutating residues and editing chains	His→Phe mutant (57 atoms), ACE–Ala–Glu–NME (37 atoms)	Replace selected residue, Delete residue (split chain), relinking two ends	B3LYP/6-31G* single point (COSMO, water)
9	Complexes of molecules and BSSE	Benzene–water dimer (15 atoms)	Placing a second molecule, Align principal axes, editing the coordinate table, BSSE (counterpoise) correction	MP2/6-31G* single point (BSSE corrected)
10	Ions and radicals	Phenol, phenoxyl radical (doublet), phenolate (charge −1)	Deleting a hydrogen, unpaired electrons and formal charges, multiplicity and reference	B3LYP/6-31+G* optimization → frequencies
11	Main-group molecules	SF₆ (octahedral), PF₅ (trigonal bipyramidal)	Non-metal coordination centers, geometries, fluoride ligands, rotational symmetry numbers	B3LYP/6-31G* optimization → frequencies
12	Solvated clusters and point charges	[Na(H₂O)₄]⁺ and TIP3P charges for an outer water	Duplicate at an offset, water to point charges, writing external point charges	B3LYP/6-31G* optimization and single points
13	Transition states and reaction paths	Transition state of F⁻ + CH₃Cl, [F···CH₃···Cl]⁻	Placing a lone atom, setting distances and angles, transition-state search, IRC and its direction	M06-2X/aug-cc-pVDZ TS search → frequencies, IRC
14	From a result to the next calculation	Optimized p-nitroaniline	History, Result viewer, Apply as candidate, excited states (TDDFT)	CAM-B3LYP/6-31+G* TDDFT
15	Isotopes and atom order	Ethanol-OD	Isotopes and masses, the coordinate table, Sort atoms	B3LYP/6-31G* frequencies
16	Your own parts (My parts)	Benzene with a trimethylsilyl group	Saving selected atoms as a part, part files, placing my parts	B3LYP/6-31G* optimization

Not in these recipes

MolForma builds finite molecules (and external point charges) and runs NWChem’s molecular methods. Not covered:

FMO (fragment molecular orbitals) and SAPT0 (energy decomposition). NWChem cannot run them, so there is no recipe that splits a protein’s active site into fragments. For reading, checking and repairing PDB files, see the “Residue” page in Right-side pages of “1 Molecule”.
Crystals and periodic systems, surfaces, automatic polymers, automatic solvent boxes, and NEB paths between two structures. Place solvent molecules by hand (recipe 12) or use a continuum model (COSMO, SMD).

An MCP client (AI) can be asked to do the same in words. The tools and examples are in “MCP server”.