The civic module

CIViCpy is primarily designed to enable exploration of the content of CIViC through Python CivicRecord objects. While these record objects can be initialized independently, the civic module also provides several routines for getting records directly from CIViC. Use of these routines is recommended.

The civic module may be imported from civicpy at the top level:

>>>from civicpy import civic

CIViC records

class civic.CivicRecord(partial=False, **kwargs)

As a base class, CivicRecord is used to define the characteristic of all records in CIViC. This class is not intended to be invoked directly by the end user, but provided for documentation of shared methods and variables in child classes.

__init__(partial=False, **kwargs)

The record object may be initialized by the user, though the practice is discouraged. To do so, values for each of the object attributes (except type) must be specified as keyword arguments, or the partial parameter must be set to True. If partial is set to True, the id keyword argument is still required.

Users are encouraged to use the functions for getting records in lieu of directly initializing record objects.

Parameters

partial (bool) – Indicates whether the the set of object attributes passed is incomplete. If set to True the id keyword is required.

type

The type of record. This field is set automatically by child classes and should not be changed.

id

The record ID. This is set on initialization using the id keyword argument, and reflects the primary ID for the record as stored in CIViC.

property site_link

Returns a URL to the record on the CIViC web application.

update(allow_partial=True, force=False, **kwargs)

Updates the record object from the cache or the server. Keyword arguments may be passed to kwargs, which will update the corresponding attributes of the CivicRecord instance.

Parameters

• allow_partial (bool) – Flag to indicate whether the record will be updated according to the contents of CACHe, without requiring all attributes to be assigned.

• force (bool) – Flag to indicate whether to force an update fromt he server, even if a full record ecists in the cache.

Returns

True if record is complete after update, else False.

CIViC record types

The primary CIViC records are found on the CIViC advanced search page, and are fully-formed

class civic.Gene(**kwargs)

Bases: civic.CivicRecord

description

A curated summary of the clinical significance of this gene.

entrez_id

The Entrez ID associated with this gene.

name

The HGNC Gene Symbol associated with this gene.

aliases

A list of alternate gene symbols by which this gene is referenced.

variants

A list of Variant records associated with this gene.

lifecycle_actions

A LifecycleAction container.

class civic.Variant(**kwargs)

Bases: civic.CivicRecord

allele_registry_id

The ClinGen Allele Registry ID associated with this variant.

civic_actionability_score

The CIViC actionability score associated with this variant.

description

A curated summary of the clinical significance of this variant.

entrez_id

The Entrez ID of the gene this variant belongs to.

entrez_name

The HGNC Gene Symbol of the gene this variant belongs to.

gene

The Gene this variant belongs to.

gene_id

The CivicRecord.id of the gene this variant belongs to.

name

The curated name given to this variant.

assertions

A list of Assertion records associated with this variant.

clinvar_entries

A list of clinvar ids associated with this variant.

coordinates

A CivicAttribute object describing CIViC coordinates.

evidence_items

evidence

A list of Evidence associated with this variant.

evidence_sources

A list of sources associated with the evidence from this variant.

hgvs_expressions

Curated HGVS expressions describing this variant.

sources

A list of sources objects associated with the variant description.

variant_aliases

aliases

A curated list of aliases by which this variant is referenced.

variant_groups

groups

A list of variant groups to which this variant belongs.

variant_types

types

A list of CivicAttribute objects describing variant types from the Sequence Ontology.

lifecycle_actions

A LifecycleAction container.

class civic.Evidence(**kwargs)

Bases: civic.CivicRecord

lifecycle_actions

A LifecycleAction container.

class civic.Assertion(partial=False, **kwargs)

Bases: civic.CivicRecord

acmg_codes

Evidence codes used in the assessment of variants under the ACMG/AMP classification guidelines.

allele_registry_id

The ClinGen Allele Registry ID associated with the assertion’s variant.

amp_level

The clinical interpretation classification by AMP/ASCO/CAP or ACMG/AMP guidelines.

clinical_significance

A string indicating the type of clinical significance statement being made, values are defined based on the corresponding evidence_type. Please see Understanding Clinical Significance for more details on the expected values for this field.

description

The Assertion Description gives detail including practice guidelines and approved tests for the variant. See curating assertions for more details.

disease

A disease CivicAttribute, linked to a corresponding Disease Ontology term when applicable.

drugs

Zero or more drug CivicAttribute, linked to corresponding NCIT terms when applicable. Only used with therapeutic response predictive evidence_type.

drug_interaction_type

One of ‘Combination’, ‘Sequential’, or ‘Substitutes’, this field describes how multiple indicated drugs within a therapeutic response predictive assertion are related.

evidence_direction

An indicator of whether the evidence statement supports or refutes the clinical significance of an event.

evidence_type

Category of clinical action/relevance implicated by event. Refer to the additional documentation on evidence types for details on how to enter evidence of each of the four types: Predictive, Prognostic, Predisposing and Diagnostic.

fda_companion_test

A boolean indicating whether or not the assertion has an associated FDA companion test.

fda_regulatory_approval

A boolean indicating whether or not the drugs indicated in the assertion have regulatory approval for use in the treatment of the assertion disease.

lifecycle_actions

A LifecycleAction container.

name

A system-generated unique identifier for the assertion, e.g. AID7.

nccn_guideline

A string linking the assertion to the corresponding NCCN Guidelines for treatment of cancer by disease site, if applicable.

nccn_guideline_version

The version associated with the indicated nccn_guideline document.

phenotypes

Zero or more phenotype CivicAttribute, linked to corresponding Human Phenotype Ontology (HPO) terms when applicable.

status

One of [‘accepted’, ‘rejected’, or ‘submitted’], describing the state of this assertion in the CIViC curation cycle.

• submitted: This assertion has been submitted by a CIViC curator or editor

• accepted: This assertion has been reviewed and approved by a CIViC editor

• rejected: This assertion has been reviewed and rejected by a CIViC editor

summary

The Assertion Summary restates the Clinical Significance as a brief single sentence statement. It is intended for potential use in clinical reports. The Assertion Summary is designed for rapid communication of the Clinical Significance, especially when displayed in a longer list with other variants.

variant

The Variant associated with this assertion.

variant_origin

The origin of this variant, one of [‘Somatic’, ‘Rare Germline’, ‘Common Germline’, ‘Unknown’, ‘N/A’, ‘Germline or Somatic’]

CIViC attributes

The CivicAttribute class is a special type of CivicRecord that is not indexed, and is used as a base class for additional complex records beyond those mentioned above (e.g. diseases, drugs). CivicAttributes are not cached except as attached objects to non-CivicAttribute CivicRecord objects, and cannot be retrieved independently.

class civic.CivicAttribute(**kwargs)

Record provenance

The LifecycleAction class is used to track the provenance of many types of CivicRecord, by serving as a container class for BaseLifecycleAction objects which in turn specify the timestamp and User associated with a given action on a record.

class civic.LifecycleAction(**kwargs)

Bases: civic.CivicAttribute

class civic.BaseLifecycleAction(**kwargs)

Bases: civic.CivicAttribute

timestamp

A CIViC timestamp string indicating the time the BaseLifecycleAction took place.

user

A CIViC User object.

class civic.Submitted(**kwargs)

Bases: civic.BaseLifecycleAction

class civic.LastModified(**kwargs)

Bases: civic.BaseLifecycleAction

class civic.LastReviewed(**kwargs)

Bases: civic.BaseLifecycleAction

class civic.Accepted(**kwargs)

Bases: civic.BaseLifecycleAction

class civic.User(**kwargs)

Bases: civic.CivicRecord

name

The user-defined full name for the User.

username

The user-defined system name for the User.

role

The CIViC role held by a User: Administrator, Editor, or Curator.

area_of_expertise

An optional attribute for a User indicating their perspective as a CIViC participant:

Research Scientist, Clinical Scientist, or Patient Advocate.

orcid

An optional attribute for a User indicating their ORCiD.

display_name

This attribute is populated with the first non-blank value from :attribute:`username`,

:attribute:`name`, :attribute:`email`, or :attribute:`id`.

created_at

A datetime describing when the User object was created.

url

A string describing a personal website or blog for the User.

twitter_handle

A string describing the Twitter handle for a User.

facebook_profile

A string describing the Facebook profile ID for a User.

linkedin_profile

A string describing the Linked profile ID for a User.

bio

A short biography for a User.

featured_expert

A flag indicating if a user is a featured expert, and thus displayed on the CIViC

domain experts section

class civic.Organization(partial=False, **kwargs)

Bases: civic.CivicRecord

name

The Organization name.

url

A URL for more information about the Organization.

description

A short text description about the Organization.

profile_image

A set of resource paths for the Organization image at varying resolution.

parent

A parent Organization, if applicable.

class civic.Country(**kwargs)

Bases: civic.CivicAttribute

iso

The ISO 3166 Country Code for the Country.

name

The full-text name for the Country.

Getting records

By ID

Records can be obtained by ID through a collection of functions provided in the civic module. Gene objects can be queried by the following methods:

civic.get_gene_by_id(gene_id)

Parameters

gene_id (int) – A single CIViC gene ID.

Returns

A Gene object.

civic.get_genes_by_ids(gene_id_list)

Parameters

gene_id_list (list) – A list of CIViC gene IDs to query against to cache and (as needed) CIViC.

Returns

A list of Gene objects.

civic.get_all_genes(include_status=['accepted', 'submitted', 'rejected'], allow_cached=True)

Queries CIViC for all genes. The cache is not considered by this function.

Parameters

• include_status (list) – A list of statuses. Only genes and their associated entities matching the given statuses will be returned.

• allow_cached (bool) – Indicates whether or not object retrieval from CACHE is allowed. If False it will query the CIViC database directly.

Returns

A list of Gene objects.

civic.get_all_gene_ids()

Queries CIViC for a list of all gene IDs. Useful for passing to get_genes_by_ids() to first check cache for any previously queried genes.

Returns

A list of all CIViC gene IDs.

Analogous methods exist for Variant, Assertion, and Evidence:

civic.get_variants_by_ids(variant_id_list)

civic.get_variant_by_id(variant_id)

civic.get_all_variants(include_status=['accepted', 'submitted', 'rejected'], allow_cached=True)

civic.get_all_variant_ids()

civic.get_assertions_by_ids(assertion_id_list=[], get_all=False)

civic.get_assertion_by_id(assertion_id)

civic.get_all_assertions(include_status=['accepted', 'submitted', 'rejected'], allow_cached=True)

civic.get_all_assertion_ids()

civic.get_all_evidence(include_status=['accepted', 'submitted', 'rejected'], allow_cached=True)

civic.get_all_evidence_ids()

By Coordinate

Variant records can be searched by GRCh37 coordinates. To query specific genomic coordinates, you will need to construct a CoordinateQuery object, and pass this query to the search_variants_by_coordinates() function. If you wish to query multiple genomic coordinates (e.g. a set of variants observed in a patient tumor), construct a sorted list of CoordinateQuery objects (sorted by chr, start, stop, alt), and pass the list to the bulk_search_variants_by_coordinates() function.

class civic.CoordinateQuery

A namedtuple with preset fields describing a genomic coordinate, for use with coordinate-based queries of CIViC Variants.

Parameters

• chr (str) – A chromosome of value 1-23, X, Y

• start (int) – The chromosomal start position in base coordinates (1-based)

• stop (int) – The chromosomal stop position in base coordinates (1-based)

• optional alt (str) – The alternate nucleotide(s) at the designated coordinates

• optional ref (str) – The reference nucleotide(s) at the designated coordinates

• build (GRCh37) – The reference build version of the coordinates

• optional key (Any) – A user-defined object linked to the coordinate

civic.search_variants_by_coordinates(coordinate_query, search_mode='any')

Search the cache for variants matching provided coordinates using the corresponding search mode.

Parameters

• coordinate_query (CoordinateQuery) – Coordinates to query

• search_mode (any,query_encompassing,variant_encompassing,exact) –

  any : any overlap between a query and a variant is a match

  query_encompassing : CIViC variant records must fit within the coordinates of the query

  record_encompassing : CIViC variant records must encompass the coordinates of the query

  exact : variants must match coordinates precisely, as well as alternate allele, if provided

  search_mode is any by default

Returns

Returns a list of variant hashes matching the coordinates and search_mode

civic.bulk_search_variants_by_coordinates(sorted_queries, search_mode='any')

An interator to search the cache for variants matching the set of sorted coordinates and yield matches corresponding to the search mode.

Parameters

• sorted_queries (list[CoordinateQuery]) – Sorted list of coordinates to query

• search_mode (any,query_encompassing,variant_encompassing,exact) –

  any : any overlap between a query and a variant is a match

  query_encompassing : CIViC variant records must fit within the coordinates of the query

  record_encompassing : CIViC variant records must encompass the coordinates of the query

  exact : variants must match coordinates precisely, as well as alternate allele, if provided

  search_mode is any by default

Returns

returns a dictionary of Match lists, keyed by query