fair-clinvar

The ClinVar database (https://www.ncbi.nlm.nih.gov/clinvar/) is a public repository of submissions from researchers on the genetic variants known in the human genome, and their assocciated diseases. The whole database can be downloaded as one gzip file in several formats, including vcf and xml. While deeply informative, this database is currently best used only on the NCBI website, and the relationships between meta-data are unclear. The database is also continually updated, (some portions daily), and the new database files are updated monthly. Therefore, we also wanted clear documentation on what we did and why. This way the method could be repeated with the new version of the database, and strengthen the arguement for changing how the database is generate/released.

This project was developed in the 2017 Bio-IT FAIR Data Hackathon.

Goals:

• Assess the FAIR qualities of the NCBI ClinVAR database according to the 15 FAIR principles
• Wrangle the database, and process using the FAIRifier (https://bioit.fair-dtls.surf-hosted.nl/fairifier/)
• Correct deficenies in the FAIRness of the database
• Create a relational scheme for the subjects (variables) in the file

Pre-processing

We found that the main vcf file contains both the whole database (over 200,000 entries and 58 columns) Since the metadata is incorporated into the file, we needed to trim the file to a proof of concept csv for FAIRizing, while including the meta-data names as header names in the tsv file.

Our initial FAIR assessment:

1. No Globally unique identifiers
2. Metadata and data in same file, therefore:
3. No metadata access when data is no longer available
4. Metadata doesn't use a broadly accessible language (assuming RDF was what was required)
5. Metadata using FAIR vocabularies - I don't think so.
6. Metadata doesn't have a complete versioning history but has some form of detailed provenance.
7. We question the, "metadata is richly described with a plurality of accurate and relevant attributes."

TSV proof-of-concept file made using python

Python script vcfparser.py

FAIRification

We submited the tsv file to the fairifier

The CLNACC field, which is RCV#, was used to make a new column for the persistent ID like https://www.ncbi.nlm.nih.gov/clinvar/RCV000148988/

Relational scheme

30,000 ft view

Using common terms

Using the metadata labels

We are working on the RDF file now.

Future directions

• Create RDF file with complete meta-data associations (~58) include stakeholder engagment
• Improve machine interoperability
• Test ML classifiers based on the relations

Self-evaluation

How well does ClinVar data align with FAIR data principles?

For each of the 15 FAIR principles, rate your dataset before and after your hackathon work. A rating of "1" is least FAIR; "5" is most FAIR.

Findable

F1. (meta)data are assigned a globally unique and persistent identifier

• Before: 2
  • Locally unique identifier.
  • Persistent URL link missing.
• After: 5
• Comment: We have assigned the RCVs as the new globally unique and persistent identifiers for the dataset

F2. data are described with rich metadata (defined by R1 below)

• Before: 4.
  • Around 60 lines of metadata within the same data file.
  • Figuring out the metadata is needs to go through the data dictionary which is multiple pages.
• After: 4
• Comment: Data has metadata which is sufficient and with a data dictionary

F3. metadata clearly and explicitly include the identifier of the data it describes

• Before: 5
  • Each line contains identifiers for diseases, syndromes, etc within the NCBI.
  • Clear linking between the metadata and data.
• After: 5
• Comment: NCBI has identifiers for diseases and syndromes with clear links between the metadata and data.

F4. (meta)data are registered or indexed in a searchable resource

• Before: 5
  • The NCBI is quite sufficient with regard to being searchable.
• After: 5
• Comment: Data is easily available on the NCBI website

Accessible

A1. (meta)data are retrievable by their identifier using a standardized communications protocol

• Before: 5
  • NCBI has identifiers based on which data can be retrieved.
• After: 5
• Comment: NCBI has retrievable methods using REST calls.

A1.1 the protocol is open, free, and universally implementable

• Before: 5
  • API info is readily available.
• After: 5
• Comment: Yes, the NCBI handles this part.

A1.2 the protocol allows for an authentication and authorization procedure, where necessary

• Before: 1
  • No authentication and authorization when required.
• After: 1
• Comment: No authentication or authorization required

A2. metadata are accessible, even when the data are no longer available

• Before: 1
  • Metadata and data are in the same file.
• After: 3
• Comment: Low utility as the data is a reported variant, closer to a concept of observation.

Interoperable

I1. (meta)data use a formal, accessible, shared, and broadly applicable language for knowledge representation.

• Before: 1
• After: 4
• Comment: We have the metadata that might not be broadly applicable but has all the other required knowledge representation

I2. (meta)data use vocabularies that follow FAIR principles

• Before: 3
  • Not globally unique locally unique though.
• After: 4
• Comment: Globally unique identifiers appended to the dataset to make it FAIR compliant

I3. (meta)data include qualified references to other (meta)data

• Before: 5
• After: 5
• Comment: NCBI provides data that’s clearly linked and easily referenced

Reusable

R1. meta(data) are richly described with a plurality of accurate and relevant attributes

• Before: 2
• After: 4
• Comment: With the RDF, the data is processed and clearly specified for reuse.

R1.1. (meta)data are released with a clear and accessible data usage license

• Before: 5
  • Public data
• After: 5

R1.2. (meta)data are associated with detailed provenance

• Before: 5
  • Dataset completely associated with version history by the NCBI
• After: 5

R1.3. (meta)data meet domain-relevant community standards

• Before: 5
  • Set up by the NCBI
• After: 5