tmqmrdfdata

tmqmrdfdata is an rdflib-based Python package designed to support and facilitate the interaction with tmQM-RDF: a Knowledge Graph Representing Transition Metal Complexes. Amongst its main functionalities, the package allows to:

• easily download the data from the dedicated GitHub repository;
• access specific subgraphs;
• retrieve the available quantitative and qualitative properties in a Python-friendly format.

Installation

The package may be installed using pip:

pip install tmqmrdfdata

Documentation

The documentation of the package is available at https://github.com/luca-cibinel/tmqmrdfdata/tree/main/docs.

Getting started

Downloading the knowledge graph

To download the tmQM-RDF knowledge graph, use the following code:

from tmqmrdfdata import download_tmQM_RDF_knowledge_graph

download_tmQM_RDF_knowledge_graph(
  dir = "data/",
  version = "latest"
)

This will download the latest available version of tmQM-RDF into the directory data/. Supposing that the latest version is version 1.0, the finaly directory tree will look like this:

data/
└── tmQM-RDF-v1.0
    ├── assertions/
    │   └── ...
    └── terminology/
        └──  ...

It is possible to download a specific version of tmQM-RDF by changing the version parameter. This parameter takes in input a string representing the exact version number to retrieve (without any leading prefix, e.g., to download the version v1.0.1, you must type version = "1.0.1").

Interfacing with the data

Once the data has been downloaded, the main interface can be instantiated:

from tmqmrdfdata import TmqmRDF

interface = TmqmRDF("data/")

This will initialise a dictionary-like object that can retrieve information from the knowledge graph. Upon instantiation, only the terminology component (or TBox) is immediately available, whereas any part of the assertion component (or ABox) will have to be explicitly retrieved first.

Accessing the TBox

The TBox contains the definition of all the terms used in the knowledge graphs. A wrapper of this part of the knowledge graph is available in the attribute interface.tbox. See the related documentation.

Accessing the ABox

Subgraphs regarding assertions on specific TMCs, ligand species, metal centres, or chemical element are retrieved using the method interface.fetch:

interface.fetch(tmcs = ["KCEYPT", "ABEVAH"], ligands = ["ligand1-0"])

Now the interface has access to the subgraphs related to the TMCs KCEYPT and ABEVAH and the ligand species ligand1-0 (according to the indexes used in tmQMg-L). These can be now accessed explicitly as follows:

kceypt = interface["TMC", "KCEYPT"]
lig1_0 = interface["ligand", "ligand1-0"]

Both kceypt and lig1-0 are instances of (subclasses) of tmqmrdfdata.assertions.TmqmRDFABoxSubgraph. Metal centres and elements can be accessed with the notation interface["centre", ...] and interface["element", ...] respectively.

Retrieval and access can be performed more conveniently using the methods .tmc, .ligand, .centre, or .element. For instance, the code

kceypt = interface.tmc("KCEYPT")
lig1_0 = interface.ligand("ligand1-0")

is equivalent to

interface.fetch(tmcs = ["KCEYPT"], ligands = ["ligand1-0"])
kceypt = interface["TMC", "KCEYPT"]
lig1_0 = interface["ligand", "ligand1-0"]

Property retrieval

Wihtin tmQM-RDF, atoms, atomic bonds, ligand species, and whole complexes are endowed with properties. These can be accessed from the corresponding TMC/ligand species subgraph. For example, if you wish to retrieve the natural atomic charge of the atoms of KCEYPT you can use the following code:

from tmqmrdfdata.terminology import tmAp

atoms_w_charge = kceypt.atoms(data = tmAp["natural_atomic_charge"])

Notice that the property had to be specified using tmAp["natural_atomic_charge"]. Let's break this symbol down:

• tmAp: this is a variable introduced in the module tmqmrdfdata.terminology. It is an rdflib.Namespace encoding the namespace <https://www.integreat.no/research/rdf/tmqm-rdf-dataset/#/atomic/atom/property/>.
• tmAp["natural_atomic_charge"] produces the URI <https://www.integreat.no/research/rdf/tmqm-rdf-dataset/#/atomic/atom/property/natural_atomic_charge>, which is the URI that tmQM-RDF uses to denote the natural atomic charge property of atoms.

If you now want to inspect the result, you will have to go through a dictionary where the keys are the URIs of the atoms of KCEYPT whereas the values are collections.namedtuple objects mirroring the structure of the RDF graph describing the property. See the related documentation for information on how this mirroring is constructed. For now, let's just inspect the first entry of this dictionary:

atom, atom_data = next(iter(atoms_w_charge.items()))

print(atom)
# >>> https://www.integreat.no/research/rdf/tmqm-rdf-dataset/#/atomic/atom/KCEYPT_Pt_0
print(atom_data.natural_atomic_charge.value)
# >>> 0.73094

Notice that, regardless of whether and which properties you request, atom_data will always have the symbol attribute, containing the chemical symbol of the atom:

print(atom_data.symbol)
# >>> https://www.integreat.no/research/rdf/tmqm-rdf-dataset/#/atomic/atom/reference/Pt

Advanced querying

If you need to perform more advanced queries, you can always rely on rdflib's own machinery. You can access the rdflib's representation of the RDF graph via the attribute .kgraph of tmqmrdfdata.assertions.TmqmRDFABoxSubgraph.

Visualising TMCs

TMC-related subgraph posses a unique method, that allows to visualise their moelcular structure using graphviz:

kceypt.view()

Contact

For any questions related to the package, contact Luca Cibinel: https://orcid.org/0009-0009-1274-8327.

For questions regarding tmQM-RDF, please check the tmQM-RDF contact info.