Skip to content

ARC Prerequisites

Knowledge Representation within the ARC

MIRA uses an ETL pipeline in the setup phase of the BrAPI server to load the content of the ARC. First the RO-Crate and then the relevant resources are read and loaded into an internal database. This database stores the research data in the BrAPI data model and provides the interface for data access by the BrAPI endpoints.

RO-Crate Ingestion

The ARC RO-Crate builds a graph describing the ARC Scaffold. Its nodes are the entities in the ARC data model which describe the experiments documented in the ARC. To extract the information the ETL Pipeline starts at the root node ./ and loads the investigation information. From there it traverses the rest of the graph following the relationships from each entity to the next.

Trial

The BrAPI Trial entity maps to the ISA Investigation. From the investigation, which is also the root element in the RO-Crate graph, the identifier, name and license are extracted. From the root element the graph is further traversed to the study entities.

Study

The ISA Study maps to the BrAPI Study entity. From the study, the identifier and name as well as the identifier and name of the investigation are extracted. The study needs to define a Growth protocol, which needs to be applied in the study's processes. Following the processes, the graph is further traversed to extract the source materials and samples.

Germplasm

The ISA Source Material maps to the BrAPI Germplasm entitiy. From the source material the following ontology annotated terms are extracted from the additionalProperties of the object of the process:

  • Genus (MIAPPE:0041)
  • Species (MIAPPE:0042)
  • Infraspecific name (MIAPPE:0043)
  • Material source DOI (MIAPPE:0051)
  • Material source latitude (MIAPPE:0052)
  • Material source longitude (MIAPPE:0053)
  • Material source coordinates uncertainty (MIAPPE:0055)
  • Material source description (MIAPPE:0056)

Observation Unit

The ISA Sample maps to the BrAPI Observation Unit entity. From the sample name and identifier are extracted from the results of the process as well as the name of the process's object and the name and identifier of the study and name and identifier of the investigation.

Trait Definitions

With MIAPPE a trait definition file describes the BrAPI Observation Variables. This file needs to be located within the corresponding study directory inside the resources directory and needs to be named tdf.tsv. It is a tab-separated values file with the headers:

  • Variable ID: Internal identifier of the observation variable
  • Variable Name: Observation variable Name
  • Variable Accession Number: PUI of the observation variable
  • Trait: Trait name
  • Trait Accession Number: PUI of the trait
  • Trait Class: Enum of morphological, agronomical, phenological, abiotic stress, biotic stress, physiological or quality
  • Method: Method name
  • Method Accession Number: PUI of the method
  • Method Description: Human readable description of the method
  • Reference associated to the method: Literature reference for the method
  • Scale: Scale name
  • Scale Accession Number: PUI of the scale
  • Scale Type: Enum of Numerical, Nominal or Ordinal
  • Scale Values: either scale unit or semicolon separated list of <code>:<value> pairs
  • Time scale: ignored

An example is given below:

./studies/myStudy/resources/tdf.tsv
1
2
3
Variable ID Variable Name   Variable Accession Number   Trait   Trait Accession Number  Trait Class Method  Method Accession Number Method Description  Reference associated to the method  Scale   Scale Accession Number  Scale Type  Scale Values    Time scale
CONVAR  Row-Type convarity      Row-Type convarity      morphological   Row-Type Estimation             1to5RowType     Nominal "1:'convar. distichon';2:'convar. deficiens';3:'convar. labile';4:'convar. intermedium';5:'convar. vulgare/convar. hexastichon'"    
AWNS_LENGTH Awns length         Awns length     morphological   Awns length Measurement             Centimeter      Numerical   cm

Observations

The actual observation data of a MIAPPE described dataset is stored in a data file, resulting of processes of executions of a Phenotyping protocol. This datafile is located with the assay directory in the dataset directory and is named df.tsv. It needs to be in long format with the headers:

  • Assay Name: Sample name
  • Date: Observation date, recommended in ISO 8601 format (YYYY-MM-DDTHH:MM:SS)
  • Trait: Variable ID of the trait definition file
  • Value: Value of the observation

An example is given below:

./assays/myAssay/dataset/df.tsv
1
2
3
Assay Name  Date    Trait   Value
HOR 4402_1105630    Tue Sep 20 00:00:00 UTC 2016    SPIKE_LENGTH    9.0
HOR 4430_1105637    Mon Sep 12 00:00:00 UTC 2016    SPIKE_LENGTH    10.0

Frequently Asked Questions

Where do I find more detailed information about MIAPPE?

The MIAPPE standard is documented within its GitHub Repository. You can find the mapping to ISA and BrAPI as well as general documentation. For a guide how to write ISA Tab, you can find information in the respective GitHub Repository.

Where do I find more information on ARC's?

You can find relevant resources for everything related to the ARC on the website. There you can find information on the Documentation Principles as well as the ARC Data Model.

Why is the data file required to be in long format?

Each observation stores a date when this observation was made. To avoid overcomplicated templates, the long format offers the simplest way of avoiding information loss.

Submit new Questions