Skip to content

Latest commit

 

History

History
145 lines (91 loc) · 4.21 KB

queries.md

File metadata and controls

145 lines (91 loc) · 4.21 KB

Proposed Query Prioritization

Summary

Here is the quick summary of the queries and their timelines.

Month 12 (T1.1.1)

  • Discovery return values

    • list of ids or URIs of relevant entities

  • Must be able to query Calls, Variants, Samples, and simple Patient data (demographics, phenotypes, diseases), and links between them

Examples:

  • Samples corresponding to patients with disease X and variants in gene Y

    • return list of sample identifiers

  • Patients with disease X and variants in gene Y that have samples of tissue Z available

    • return list of patient identifiers

Month 18 (T1.1.2)

  • More complex return values

    • histogram of counts, min/max, etc

  • Include RNA expression

  • Deeper clin/phen fields, in collaboration w/ WP3

    • include capabilities for treatments, outcomes, lab work, references to other data

Examples:

  • Hypothesis-driven association query: contingency table of phenotype X vs mutations in gene Y

  • Patients with disease X and mutations in gene Y with RNA expression data available in gene Y

  • Histogram of variants in pathway X in patients with phenotype Y by counts; or by consequence

  • Counting of patients in registry with a specific rare disease

  • Information retrevial queries on some fields (or push to D1.2, Month 24?)

    • need good examples for this
    • is this part of the underlying query API or built on top?

A longer description of the entities, queries, prioritization, and rationale follows. For more informatino, consult the list of collected use cases.

Entities

The entities (and their fields) we aim to query over the next 18 months as part of CINECA WP1 are identified as:

Genomics

  • Variants, with annotations

    • Gene, Pathway
    • Type, Consequence
    • Germline/Somatic (sample)

  • Calls

    • Variants
    • Samples/Patients
    • Interpretation

  • RNA Expression

    • Samples associated if any
    • Patients associated if any

Specimen

  • Samples
    • Tissue Type
    • Genomic data, Patients associated if any

Clinical/Phenotypic

  • Patients
    • Demographic info
    • Phenotypes
    • Diseases
    • Treatments
    • Outcomes
    • Samples associated if any
    • Genomic data associated if any
    • Lab work
    • Other data modalities (links to imaging, etc)

Requirements/Constraints

Hard requirements for query implementations include:

  • Clin/Phen + Variants needed for short term WP5 use cases
  • Clin/Phen + Samples needed for biobanking use cases
  • Not all data sets will have all fields; handle this gracefully
  • Must be able to make use of existing infrastructure
  • Biobanking use cases will also have needs for information-retrieval type queries

Queries

For Month 12 (Dec 2019)

  • Discovery return values

    • list of ids or URIs of relevant entities

  • No RNA expression (will be easier later when GA4GH RNA expression data formats/APIs well established)

  • Must be able to query Calls, Variants, Samples, and simple Patient data (demographics, phenotypes, diseases), and links between them

Examples:

  • Samples corresponding to patients with disease X and variants in gene Y

    • return list of sample identifiers

  • Patients with disease X and variants in gene Y that have samples of tissue Z available

    • return list of patient identifiers

For Month 18 (Jun 2020)

  • Add Treatments, Outcomes, Lab work, Other data modalities to clin/phen

  • Add access to RNA expression data

    • GA4GH API/standard matrix format should be well established by then

  • Add information retrieval queries (fuzzy search)

    • Need concrete examples

  • Add complex return values

    • e.g., List/Histogram of Patient.Outcomes where variant X is present and treatment Y was undergone

Examples:

  • Hypothesis-driven association query: contingency table of phenotype X vs mutations in gene Y

  • Patients with disease X and mutations in gene Y with RNA expression data available in gene Y

  • Histogram of variants in pathway X in patients with phenotype Y by counts; or by consequence

  • Counting of patients in registry with a specific rare disease

  • Information retrevial queries on some fields (or push to D1.2, Month 24?)

    • need good examples for this
    • is this part of the underlying query API or built on top?