Molecules are searched from the number specified by "Start PCCDB-ID" as the starting point.
The number specified by "Limit" gives the upper limit of molecular data (search results) in the query.
At present, the maximize number of "Limit" is 2,000 molecules.
(This restriction will be relaxed by sever augmentations in the near future.)
To set a condition, it is necessary to mark a checkbox.
The input forms followed by each checkbox indicate "lower" and "upper" bounds for the specified property.
Here, the range of HOMO-LUMO gap and energies are specified in eV.
If the lower (upper) form is not input in the submission, no lower (upper) bound is set in the search.
The "Oscillator strength" entry can be used to specify the photon absorption of the excited state.
To search molecules containing specified elements, please check the "Select elements" checkbox, and choose target elements.
At that time, if a element is unchecked, our system searches molecules which do not contain the element.
If "String match" is checked, molecules are searched based on the SMILES representations registered in PubChem.
The SMILES representation may not be uniquely determined even for the same molecular (sub)structure.
To find molecules, the same SMILES representation used in PubChem may give better results in molecule searches.
For string matching, we convert the inputed SMILES representation to an internal representation, if possible.
Thus, the input string is not always used as it is. While the string matching is strict, and has many restrictions,
the similarity search may have more flexibility to find molecules.
PubChem provides a service to create the SMILES representation from the 2D (sub)structure of molecules.
It may help your searches.
By checking "Similarity search", molecules can be searched based on the Tanimoto similarity with the molecular fingerprint techniques.
The MACCS (166 bits) and Topological (1024 bits) fingerprints are available, which are implemented in Open Babel
and RDkit, respectively.
If the "both" radio button is checked, fingerprints that concatenate those are used.
If no value is specified for "threshold", a default value of 0.3 is used in the similarity search.
The similarity search takes a lot of sever resources. Please use it carefully.
The number of hydrogen bond acceptor/donor (HBA/HBD), TPSA, logP, molecular refractivity (MR), and melting point (MP) were calculated
by using Open Babel.