Q1: Do formal charge assignment and
stereochemistry enumeration need to be done prior to running
GINGER to generate conformers (.molt file)?
A:
Not necessarily — it depends on your input and goals:
Formal Charges:
If your input molecules already have formal charges assigned
and you want to preserve them, no additional steps are needed.
If the charges are missing or incomplete, use the -Cn option (see Q2
below) to have GINGER assign formal charges automatically.
Stereochemistry:
GINGER preserves defined stereocenters as-is. For undefined
(racemic) centers, GINGER samples stereoisomers during
conformer generation (see Q3 below). However, the number of
conformers per stereoisomer will be limited, since the total
number of conformers is divided across all possible
stereoisomers.
For better coverage — especially for molecules with multiple
racemic centers — it's recommended to enumerate stereoisomers
beforehand. This way, GINGER will generate a full set of
conformers for each defined stereoisomer individually.
Q2: Can GINGER assign charges automatically, and
what models are available?
A:
Yes. GINGER can assign formal charges automatically based on
predicted protonation states:
Use the -C option to apply MolSoft’s standard pKa-based model.
Use the -Cn option (in newer versions) to apply a neural network-assisted pKa model, which may provide improved accuracy in some cases.
Q3: What is the behavior of GINGER when a
racemic chiral center is present? Are N conformers generated per
enantiomer or in total?
A:
GINGER will generate up to N conformers total. For each conformer,
random stereochemistry is assigned at racemic centers.
For example, if there is one racemic center and N = 30, the output
typically includes approximately 15 conformers for each
enantiomer, depending on sampling randomness.
Q4: How does GINGER handle double bonds when no
cis/trans specification is given?
A:
If the input does not restrict double bond configuration (e.g.,
unspecified in SMILES), GINGER will generate a mixture of cis and
trans conformers within the total N limit.
The distribution depends on the chemical environment and training
statistics — sometimes near-equal, sometimes skewed toward one
isomer.