ROBOT

- - - - - - - - - -
view on github
getting started
common errors
chaining commands
global options
makefile
plugins
- - - - - - - - - -
annotate
collapse
convert
diff
expand
explain
export
export-prefixes
extract
filter
materialize
measure
merge
mirror
python
query
reason
reduce
relax
remove
rename
repair
report
template
unmerge
validate-profile
verify
- - - - - - - - - -
ROBOT is licensed under the
BSD 3-Clause License.
Theme by orderedlist

Extract

Contents

  1. Overview
  2. Extraction method: SLME
  3. Extraction method: MIREOT
  4. Extraction method: Subset
  5. Handling Imports (--imports)
  6. Extracting Ontology Annotations (--copy-ontology-annotations)
  7. Adding Source Annotations (--annotate-with-source)

Overview

The reuse of ontology terms creates links between data, making the ontology and the data more valuable. But often you want to reuse just a subset of terms from a target ontology, not the whole thing. Here we extract a “STAR” module for the term ‘adrenal cortex’ and its supporting terms:

robot extract --method STAR \
    --input filtered.owl \
    --term-file uberon_module.txt \
    --output results/uberon_module.owl

See uberon_module.txt for an example of a term file. Terms should be listed line by line in either CURIE form or full IRI, and comments can be included with one or more whitespace characters after the term followed by # and then the comment.

Alternatively, individual terms can be specified with --term followed by the CURIE (or full URI).

The --method options fall into two groups: Syntactic Locality Module Extractor (SLME) and Minimum Information to Reference an External Ontology Term (MIREOT).

Syntactic Locality Module Extractor (SLME)

Each SLME module type takes a “seed” that you specify with --term and --term-file options. From the seed it builds a module with a “signature” that includes the seed plus any other terms required so that any logical entailments are preserved between entities (classes, properties and individuals) in the signature. For example, if an ontology implies that A is a subclass of B, and the seed contains A and B, then the module will also imply that A is a subclass of B. In other words, the module will contain all the axioms needed to provide the same entailments for the seed terms (and resulting signature) as the full ontology would.

For more details see:

ROBOT expects any --term or IRI in the --term-file to exist in the input ontology. If none of the input terms exist, the command will fail with an empty terms error. This can be overridden by including --force true.

Instances

Important note for ontologies that include individuals: When using the SLME method of extraction, all individuals (ABox axioms) and their class types (the TBox axioms they depend on) are included by default. The extract command provides an --individuals option to specify what (if any) individuals are included in the output ontology:

MIREOT

The MIREOT method preserves the hierarchy of the input ontology (subclass and subproperty relationships), but does not try to preserve the full set of logical entailments. Both “upper” (ancestor) and “lower” (descendant) limits can be specified, like this:

robot extract --method MIREOT \
    --input uberon_fragment.owl \
    --upper-term "obo:UBERON_0000465" \
    --lower-term "obo:UBERON_0001017" \
    --lower-term "obo:UBERON_0002369" \
    --output results/uberon_mireot.owl

To specify upper and lower term files, use --upper-terms and --lower-terms. The upper terms are the upper boundaries of what will be extracted. If no upper term is specified, all terms up to the root (owl:Thing) will be returned. The lower term (or terms) is required; this is the limit to what will be extracted, e.g. no descendants of the lower term will be included in the result.

To only include all descendants of a term or set of terms, use --branch-from-term or --branch-from-terms, respectively. --lower-term or --lower-terms are not required when using this option.

Note that if the same IRI is used for both a class and an individual, MIREOT will ignore the individual and only extract the class.

For more details see the MIREOT paper.

Subset

The subset method extracts a sub-ontology that contains only the seed terms (that you specify with --term and --term-file options) and the relations between them. This method uses the relation-graph to materialize the existential relations among the seed terms. Procedurally, the subset method materializes the input ontology and adds the inferred axioms to the input ontology. Then filters the ontology with the given seed terms. Finally, it reduces the filtered ontology to remove redundant subClassOf axioms.

robot extract --method subset \
    --input subset.obo \
    --term "obo:ONT_1" \
    --term "obo:ONT_5" \
    --term "BFO:0000050" \
    --output results/subset_result.owl

ROBOT expects any --term or IRI in the --term-file to exist in the input ontology. If none of the input terms exist, the command will fail with an empty terms error. This can be overridden by including --force true.

Intermediates

When extracting (especially with MIREOT), sometimes the hierarchy can have too many intermediate classes, making it difficult to identify relevant relationships. For example, you may end up with this after extracting adrenal gland:

- material anatomical entity
  - anatomical structure
    - multicellular anatomical structure
      - organ
        - abdomen element
          - adrenal/interrenal gland
            - adrenal gland (*)
    - lateral structure
      - adrenal gland (*)

By specifying how to handle these intermediates, you can reduce unnecessary intermediate classes:

Running this command to extract, inclusively, between ‘material anatomical entity’ and ‘adrenal gland’:

robot extract --method MIREOT \
    --input uberon_fragment.owl \
    --upper-term UBERON:0000465 \
    --lower-term UBERON:0002369 \
    --intermediates minimal \
    --output results/uberon_minimal.owl

Would result in the following structure:

- material anatomical entity
  - anatomical structure
    - adrenal gland (*)
    - organ
      - adrenal gland (*)

You can chain this output into reduce to further clean up the structure, as some redundant axioms may appear.

Running the same command, but with --intermediates none:

robot extract --method MIREOT \
    --input uberon_fragment.owl \
    --upper-term UBERON:0000465 \
    --lower-term UBERON:0002369 \
    --intermediates none \
    --output results/uberon_none.owl

Would result in:

- material anatomical entity
  - adrenal gland

Any term specified as an input term will not be pruned.

Handling Imports

By default, extract will include imported ontologies. To exclude imported ontologies, just add --imports exclude for any non-MIREOT extraction method:

robot extract --method BOT \
  --catalog catalog.xml \
  --input imports-nucleus.owl \
  --term GO:0005739 \
  --imports exclude \
  --output results/mitochondrion.owl

This only includes what is asserted in imports-nucleus.owl, which imports nucleus.owl. imports-nucleus.owl only includes the term ‘mitochondrion’ (GO:0005739) and links it to its parent class, ‘intracellular membrane-bounded organelle’ (GO:0043231). nucleus.owl contains the full hierarchy down to ‘intracellular membrane-bounded organelle’. The output module, mitochondrion.owl, only includes the term ‘mitochondrion’ and this subClassOf statement.

By contrast, including imports returns the full hierarchy down to ‘mitochondrion’, which is asserted in nucleus.owl:

robot extract --method BOT \
  --catalog catalog.xml \
  --input imports-nucleus.owl \
  --term GO:0005739 \
  --imports include \
  --output results/mitochondrion-full.owl

Extracting Ontology Annotations

You can also include ontology annotations from the input ontology with --copy-ontology-annotations true. By default, this is false.

robot extract --method BOT \
  --input annotated.owl \
  --term UBERON:0000916 \
  --copy-ontology-annotations true \
  --output results/annotated_module.owl

Adding Source Annotations

extract provides an option to annotate extracted terms with rdfs:isDefinedBy. If the term already has an annotation using this property, the existing annotation will be copied and no new annotation will be added.

robot extract --method BOT \
  --input annotated.owl \
  --term UBERON:0000916 \
  --annotate-with-source true \
  --output results/annotated_source.owl 

The object of the property is, by default, the base name of the term’s IRI. For example, the IRI for GO:0000001 (http://purl.obolibrary.org/obo/GO_0000001) would receive the source http://purl.obolibrary.org/obo/go.owl.

Sometimes classes are adopted by other ontologies, but retain their original IRI. In this case, you can provide the path to a term-to-source mapping file as CSV or TSV.

robot --prefix 'GO: http://purl.obolibrary.org/obo/GO_' \
  extract --method BOT \
  --input annotated.owl \
  --term UBERON:0000916 \
  --annotate-with-source true \
  --sources source-map.tsv \
  --output results/changed_source.owl

The mapping file can either use full IRIs:

http://purl.obolibrary.org/obo/BFO_0000001,http://purl.obolibrary.org/obo/ro.owl

Or prefixes, as long as the prefix is valid:

BFO:0000001,RO

Error Messages

Missing MIREOT Term(s) Error

MIREOT requires either --lower-term or --branch-from-term to proceed. --upper-term is optional.

Missing Lower Term(s) Error

If an --upper-term is specified for MIREOT, --lower-term (or terms) must also be specified.

Invalid Imports Error

The input for --imports must be either exclude or include.

Invalid Method Error

The --method option only accepts: MIREOT, STAR, TOP, and BOT.

Invalid Option Error

The following flags should not be used with STAR, TOP, or BOT methods:

Invalid Source Map Error

The input for --sources must be either CSV or TSV format.

Unknown Individuals Error

--individuals must be one of: include, minimal, definitions, or exclude.

Unknown Intermediates Error

--intermediates must be one of: all, minimal, or none.