biointerface package

Submodules

biointerface.core module

Core module for extracting Protein-nucleic acid interfaces.

class biointerface.core.Interface(contacts: list[tuple[Atom, Atom]], search_radius: int | float = 4.0)[source]

Bases: object

Class for Protein-nucleic acid interface.

Parameters

contactslist[tuple[Atom, Atom]]

List of pairs of atoms, first one is from the nucleic acids, second one is from the protein.

search_radiusfloat | int, optional

Search radius, measured in Angstrom, within which Protein-nucleic acid interactions are found. Default is 4.0

as_dataframe() → DataFrame[source]

Get all data from the interface, as a dataframe.

Contains the following data fields:

Residue hetero field Residue number Residue insertion code Residue name Atom name Atom alternate location Atom element Atomic coordinates (x, y, z) From both protein and nucleic acid atoms Euclidean distance between atom pair in contact

Returns

dfpd.DataFrame

All data from the interface.

get_aminoacids() → list[Residue][source]

Get only protein residues in the protein-nucleic acid interface.

Returns

list[Residue]

List of protein residues in the interface.

get_atomic_contacts() → list[tuple[Atom, Atom]][source]

Get interface contacts as pairs of atoms.

Returns

list[tuple[Atom, Atom]]

List of pairs of atoms, first one is from the nucleic acids, second one is from the protein.

get_binding_domain() → Polypeptide | None[source]

Get nucleic acid binding domain from the protein.

The output is the binding “gapped” subsequence of the full protein found in the structure.

This method allows for “gaps” of unbound aminoacids inside the binding domain, only the aminoacids at the ends are trimmed according to being bound to nucleic acids (NAs) or not.

A visual example of “gaps”:

` Input full protein:          MQMLLNHKPTKFNGAIDERFHWKVIQRISGSEG NA-bound:                               ****  ** Output binding domain:                  FNGAIDER `

This method is only an inference of the NA-binding domain: while the output will likely align with the annotated true domain, it’ll likely not infer the whole domain. This is because a domain is defined by folding properties, while this method is much more naive. This is why I implemented some “padding” on both ends of the binding domain, it allows to be more lenient of the extent of the binding domain.

Returns

binding_domainPolypeptide | None

Nucleic acid binding domain.

get_binding_protein() → Polypeptide | None[source]
get_bound_double_strands() → list[DoubleStrandNucleicAcid] | None[source]

Get all double strand nucleic acids bound by the protein.

Returns

list[DoubleStrandNucleicAcid] | None

List of double strand nucleic acids bound by the protein.

get_bound_nucleic_acids() → list[NucleicAcid] | None[source]

Get all nucleic acids bound by the protein.

Returns

list[NucleicAcid] | None

List of nucleic acids bound by the protein.

get_nucleic_acid_atoms() → list[Atom][source]

Get only nucleic acid atoms in the protein-nucleic acid interface.

Returns

list[Atom]

List of nucleic acid atoms in the interface.

get_nucleotides() → list[Residue][source]

Get only nucleic acid residues in the protein-nucleic acid interface.

Returns

list[Residue]

List of nucleic acid residues in the interface.

get_protein_atoms() → list[Atom][source]

Get only protein atoms in the protein-nucleic acid interface.

Returns

list[Atom]

List of protein atoms in the interface.

get_trimmed_double_strands() → list[DoubleStrandNucleicAcid] | None[source]

Get all double strand nucleic acids bound by the protein, but trimmed by binding.

The output double strand nucleic acids (NAs) are subsequences of the full DSNAs found in the structure, since proteins might not bind the whole DSNA.

This method allows for “gaps” of unbound base pairs inside the DSNA, only the base pairs at the ends are trimmed according to being protein-bound or not.

A visual example of “gaps”:

``` Input full DSNA: GATATACAAGCCA

TGGCTTGTATATC

Protein-bound: **** ** Output protein-bound DSNA: TATACAAG

CTTGTATA

```

Returns

list[DoubleStrandNucleicAcid] | None

List of double strand nucleic acids bound by the protein, but trimmed by binding.

get_trimmed_nucleic_acids() → list[NucleicAcid] | None[source]

Get all nucleic acids bound by the protein, but trimmed by binding.

The output nucleic acids (NAs) are subsequences of the full NAs found in the structure, since proteins might not bind the whole NA.

This method allows for “gaps” of unbound nucleotides inside the NA, only the nucleotides at the ends are trimmed according to being protein-bound or not.

A visual example of “gaps”:

` Input full NA:            GATATACAAGCCA Protein-bound:              ****  ** Output protein-bound NA:    TATACAAG `

Returns

list[NucleicAcid] | None

List of nucleic acids bound by the protein, but trimmed by binding.

class biointerface.core.InterfaceBuilder(search_radius: float | int = 4.0)[source]

Bases: object

Use atomic distance to find Protein-Nucleic acid interfaces.

Assuming you only want standard nucleotides and amino acids.

Parameters

search_radiusfloat | int, optional

Search radius, measured in Angstrom, within which Protein-Nucleic acid interactions are found. Default is 4.0

build_interfaces(entity: Structure | Model | Chain, build_by_protein_chain: bool = True, pp_builder: PPBuilder = <Bio.PDB.Polypeptide.PPBuilder object>, standard_aminoacids: bool = True, na_builder: NABuilder = <PDBNucleicAcids.NucleicAcid.NABuilder object>, dsna_builder: DSNABuilder = <PDBNucleicAcids.NucleicAcid.DSNABuilder object>, standard_nucleotides: bool = True, pairing_rules: BasePairRules = <PDBNucleicAcids.BasePairRules.WatsonCrickBasePairRules object>) → list[Interface][source]

Extract all Protein-Nucleic acid interfaces found in a PDB entity.

Parameters

entityL{Structure}, L{Model} or L{Chain}

Protein-nucleic acid interfaces are searched for in this object. L{Structure} is the suggested input.

build_by_protein_chain: bool, optional

If True, builds interfaces not by polypeptide binding but by chain binding: multiple polypeptides coming from a single chain will be fused together into a single unconnected NA-binding polypeptide. Default is True.

pp_builderPPBuilder, optional

Polypeptide builder class from Biopython. Default is PPBuilder with default parameters.

standard_aminoacids: bool, optional

Use only standard aminoacids. This is the aa_only parameter in the PPBuilder.build_peptides() method. Default is True.

na_builderNABuilder, optional

Polypeptide builder class from PDBNucleicAcids. Default is NABuilder with default parameters.

dsna_builderDSNABuilder, optional

Polypeptide builder class from PDBNucleicAcids. Default is DSNABuilder with default parameters.

standard_nucleotides: bool, optional

Use only standard nucleotides. This parameter is used in the NABuilder.build_nucleic_acids() method and in the DSNABuilder.build_double_strands() method. Default is True.

pairing_rulesoptional

Rules for proper base pairing class instance from PDBNucleicAcids. This parameter is used in the DSNABuilder.build_double_strands() method. Default is WatsonCrickBasePairRules() with default parameters.

Raises

PDBConstructionException: In case there is no protein

in the input entity.

PDBConstructionException: In case there is no nucleic acid

in the input entity.

Returns

list[Interface]

List of all Protein-Nucleic acid interfaces found in a PDB entity.

Module contents

Top-level package for BioInterface.

class biointerface.Interface(contacts: list[tuple[Atom, Atom]], search_radius: int | float = 4.0)[source]

Bases: object

Class for Protein-nucleic acid interface.

Parameters

contactslist[tuple[Atom, Atom]]

List of pairs of atoms, first one is from the nucleic acids, second one is from the protein.

search_radiusfloat | int, optional

Search radius, measured in Angstrom, within which Protein-nucleic acid interactions are found. Default is 4.0

as_dataframe() → DataFrame[source]

Get all data from the interface, as a dataframe.

Contains the following data fields:

Residue hetero field Residue number Residue insertion code Residue name Atom name Atom alternate location Atom element Atomic coordinates (x, y, z) From both protein and nucleic acid atoms Euclidean distance between atom pair in contact

Returns

dfpd.DataFrame

All data from the interface.

get_aminoacids() → list[Residue][source]

Get only protein residues in the protein-nucleic acid interface.

Returns

list[Residue]

List of protein residues in the interface.

get_atomic_contacts() → list[tuple[Atom, Atom]][source]

Get interface contacts as pairs of atoms.

Returns

list[tuple[Atom, Atom]]

List of pairs of atoms, first one is from the nucleic acids, second one is from the protein.

get_binding_domain() → Polypeptide | None[source]

Get nucleic acid binding domain from the protein.

The output is the binding “gapped” subsequence of the full protein found in the structure.

This method allows for “gaps” of unbound aminoacids inside the binding domain, only the aminoacids at the ends are trimmed according to being bound to nucleic acids (NAs) or not.

A visual example of “gaps”:

` Input full protein:          MQMLLNHKPTKFNGAIDERFHWKVIQRISGSEG NA-bound:                               ****  ** Output binding domain:                  FNGAIDER `

This method is only an inference of the NA-binding domain: while the output will likely align with the annotated true domain, it’ll likely not infer the whole domain. This is because a domain is defined by folding properties, while this method is much more naive. This is why I implemented some “padding” on both ends of the binding domain, it allows to be more lenient of the extent of the binding domain.

Returns

binding_domainPolypeptide | None

Nucleic acid binding domain.

get_binding_protein() → Polypeptide | None[source]
get_bound_double_strands() → list[DoubleStrandNucleicAcid] | None[source]

Get all double strand nucleic acids bound by the protein.

Returns

list[DoubleStrandNucleicAcid] | None

List of double strand nucleic acids bound by the protein.

get_bound_nucleic_acids() → list[NucleicAcid] | None[source]

Get all nucleic acids bound by the protein.

Returns

list[NucleicAcid] | None

List of nucleic acids bound by the protein.

get_nucleic_acid_atoms() → list[Atom][source]

Get only nucleic acid atoms in the protein-nucleic acid interface.

Returns

list[Atom]

List of nucleic acid atoms in the interface.

get_nucleotides() → list[Residue][source]

Get only nucleic acid residues in the protein-nucleic acid interface.

Returns

list[Residue]

List of nucleic acid residues in the interface.

get_protein_atoms() → list[Atom][source]

Get only protein atoms in the protein-nucleic acid interface.

Returns

list[Atom]

List of protein atoms in the interface.

get_trimmed_double_strands() → list[DoubleStrandNucleicAcid] | None[source]

Get all double strand nucleic acids bound by the protein, but trimmed by binding.

The output double strand nucleic acids (NAs) are subsequences of the full DSNAs found in the structure, since proteins might not bind the whole DSNA.

This method allows for “gaps” of unbound base pairs inside the DSNA, only the base pairs at the ends are trimmed according to being protein-bound or not.

A visual example of “gaps”:

``` Input full DSNA: GATATACAAGCCA

TGGCTTGTATATC

Protein-bound: **** ** Output protein-bound DSNA: TATACAAG

CTTGTATA

```

Returns

list[DoubleStrandNucleicAcid] | None

List of double strand nucleic acids bound by the protein, but trimmed by binding.

get_trimmed_nucleic_acids() → list[NucleicAcid] | None[source]

Get all nucleic acids bound by the protein, but trimmed by binding.

The output nucleic acids (NAs) are subsequences of the full NAs found in the structure, since proteins might not bind the whole NA.

This method allows for “gaps” of unbound nucleotides inside the NA, only the nucleotides at the ends are trimmed according to being protein-bound or not.

A visual example of “gaps”:

` Input full NA:            GATATACAAGCCA Protein-bound:              ****  ** Output protein-bound NA:    TATACAAG `

Returns

list[NucleicAcid] | None

List of nucleic acids bound by the protein, but trimmed by binding.

class biointerface.InterfaceBuilder(search_radius: float | int = 4.0)[source]

Bases: object

Use atomic distance to find Protein-Nucleic acid interfaces.

Assuming you only want standard nucleotides and amino acids.

Parameters

search_radiusfloat | int, optional

Search radius, measured in Angstrom, within which Protein-Nucleic acid interactions are found. Default is 4.0

build_interfaces(entity: Structure | Model | Chain, build_by_protein_chain: bool = True, pp_builder: PPBuilder = <Bio.PDB.Polypeptide.PPBuilder object>, standard_aminoacids: bool = True, na_builder: NABuilder = <PDBNucleicAcids.NucleicAcid.NABuilder object>, dsna_builder: DSNABuilder = <PDBNucleicAcids.NucleicAcid.DSNABuilder object>, standard_nucleotides: bool = True, pairing_rules: BasePairRules = <PDBNucleicAcids.BasePairRules.WatsonCrickBasePairRules object>) → list[Interface][source]

Extract all Protein-Nucleic acid interfaces found in a PDB entity.

Parameters

entityL{Structure}, L{Model} or L{Chain}

Protein-nucleic acid interfaces are searched for in this object. L{Structure} is the suggested input.

build_by_protein_chain: bool, optional

If True, builds interfaces not by polypeptide binding but by chain binding: multiple polypeptides coming from a single chain will be fused together into a single unconnected NA-binding polypeptide. Default is True.

pp_builderPPBuilder, optional

Polypeptide builder class from Biopython. Default is PPBuilder with default parameters.

standard_aminoacids: bool, optional

Use only standard aminoacids. This is the aa_only parameter in the PPBuilder.build_peptides() method. Default is True.

na_builderNABuilder, optional

Polypeptide builder class from PDBNucleicAcids. Default is NABuilder with default parameters.

dsna_builderDSNABuilder, optional

Polypeptide builder class from PDBNucleicAcids. Default is DSNABuilder with default parameters.

standard_nucleotides: bool, optional

Use only standard nucleotides. This parameter is used in the NABuilder.build_nucleic_acids() method and in the DSNABuilder.build_double_strands() method. Default is True.

pairing_rulesoptional

Rules for proper base pairing class instance from PDBNucleicAcids. This parameter is used in the DSNABuilder.build_double_strands() method. Default is WatsonCrickBasePairRules() with default parameters.

Raises

PDBConstructionException: In case there is no protein

in the input entity.

PDBConstructionException: In case there is no nucleic acid

in the input entity.

Returns

list[Interface]

List of all Protein-Nucleic acid interfaces found in a PDB entity.