An Open Repository Model for Acquiring Knowledge about Scientific Experiments

EKAW 2016 – November 21th, 2016Bologna, Italy

Martin O’Connor, Marcos Martínez-Romero, Attila L. Egyedi, Debra Willrett, John Graybeal, and Mark A. Musen

Stanford University, Stanford, CA, USA

Stanford University metadatacenter.org

Reproducibility Problem in Science

Metadata Key to Addressing Problem

•  Crucial for reproducibility in biomedicine–  Locate experimental datasets online–  Understand how the experiments were performed–  Reuse the data to perform new analyses

•  Journals and funding agencies increasingly require making experimental data and metadata available

Many Metadata Standards have been Developed

However: Metadata Submission is Hard

Metadata  

Summary  Data  Matrix  

Raw  Data  

Submission  Interface  

Metadata Submission is Hard - II

age Age AGE `Age

age (after birth) age (in years)

age (y) age (year) age (years) Age (years) Age (Years)

age (yr) age (yr-old)

age (yrs) Age (yrs)

age [y] age [year] age [years] age in years

age of patient Age of patient age of subjects

age(years) Age(years) Age(yrs.) Age, year age, years

age, yrs age.year

age_years

Result: Poor Metadata

Variants  of  ‘age’  metadata  field  in  Gene  Expression  Omnibus  (GEO)  repository  

Our Solution: CEDAR - A Metadata Ecosystem

•  Overcome the impediments to creating high-quality metadata

•  Facilitate–  Creation–  Acquisition–  Use–  Evaluation–  Refinement

•  Key goal: create a sharable metadata exchange format – a template model - for publishing, searching, exchanging metadata

CEDAR Template Model Goals•  Must describe composite

structure of templates•  Implemented using standard

formats•  Express semantics•  Metadata instances:

–  Linked to controlled terms–  Easily serializable–  Easily validated–  Easily indexed–  Interchange with RDF–  High readable–  Produced/consumed via

REST APIs and usable in JavaScript front ends

–  Meets FAIR goals

Study

Principal Investigator

Description

Name

Institution

Name

ZIP

Title

Metadata Template

FieldsTemplateElements

JSON Schema + JSON-LD JSON-LD

Using JSON Schema and JSON-LD for CEDAR Template Model

What is JSON Schema?•  Technology for describing and validating the

structure of JSON documents

•  Provides a structural description of any JSON document

•  JSON documents that are specified with JSON Schema can be structurally validated against their associated schemas

•  Analogous to XML Schema

What is JSON-LD?•  A lightweight syntax to serialize Linked Data in JSON

•  Allows existing JSON to be interpreted as Linked Data with minimal changes

•  JSON-LD is primarily intended to be a way to:–  use Linked Data in Web-based programming environments–  build interoperable Web services–  store Linked Data in JSON-based storage engines

•  Core contribution: add semantics to JSON documents

•  W3C Recommendation: https://www.w3.org/TR/json-ld/

{ "$schema": "http://json-schema.org/draft-04/schema#", "@type": "https://repo.metadatacenter.org/core/Template",

"@id": "https://repo.metadatacenter.org/templates/434334", "title": ”Study", "description": ”Study template", "type": "object", "_ui": {...}, "properties": {

"title": {...}, ”description": {...}, ”principalInvestigator": {...} }, "required": ["title", "description", "principalInvestigator"]

"additionalProperties": false}

Using JSON Schema to Define Template Structure

{ "title": { "@value": "Immune biomarkers study" }, "description": { "@value": "Immune biomarkers …" }, "principalInvestigator": {

"name": { "@value": "Dr. P.I" }, "institution": { "name": { "@value": "Stanford" },

"zip": { "@value": "94305" } } }}

Using JSON-LD to add Semantics to Metadata Instances

{ "@type": "http://semantic-dicom.org/dcm#Study", "@id": "https://repo.metadatacenter.org/template_instances/55417", "@context": {

"title": "https://schema.org/title", "name": "https://schema.org/name", "description": "https://schema.org/description",

"zip": "https://schema.org/postalCode", "pi": "https://myschema.org/property/hasPI", "institution": "https://myschema.org/property/hasInstitution" },

"title": { "@value": "Immune biomarkers study" }, "description": { "@value": "Immune biomarkers …" }, ”principalInvestigator": {

"@type": "https://schema.org/Person", "@id": "https://repo.metadatacenter.org/template_elements/557", "name": { "@value": "Dr. P.I" }, "institution": {

"@type": "https://schema.org/Organization", "@id": "https://repo.metadatacenter.org/template_elements/37", "name": { "@value": "Stanford" }, "zip": { "@value": "94305" }

} }}

Using JSON-LD to add Semantics to Metadata Instances - II

CEDAR Metadata Instances can be transformed to an RDF Graph

tinstances:55417

telements:37

telements:557

Immune biomarkers study

Immune biomarkers …

schema:Organization

schema:Person

dcm:Study

Dr. P.I.

Stanford

94305

rdf:type

rdf:type

rdf:type

schema:nameschema:description

schema:name

schema:nameschema:postalCode

myschema:hasPI

myschema:hasInstitution

CEDAR  Template  Model  

Controlled  terminologies  

Model drives CEDAR Workbench

Template Designer provides Template Creation

Metadata Editor automatically generates an Acquisition Interface

Metadata Editor Adds Semantics

Initial Results•  Public alpha release in September 2016•  Represented all public metadata in

ImmPort repository (146 studies)•  Represented an array of public ISA-

created biomedical studies (~300)•  Represented 60k ISO 11179-based

Common Data Elements from NCI•  Currently working with Stanford Digital

Repository and several research groups

Summary•  We have developed a standards-based

template model for representing, publishing, and sharing templates and metadata

•  Provides strong interoperation with Linked Open Data

•  Metadata easy to create/consume using off-the-shelf tools

•  Very easy to work with using CEDAR tools

CEDAR Resources•  Web site: http://metadatacenter.org•  Workbench: https://cedar.metadatacenter.net•  GitHub: https://metadatacenter.github.io