JPhyloRef

JPhyloRef is a Java-based command line tool as well as a web service for reasoning with OWL ontologies containing phyloreferences and their accompanying reference phylogenetic trees. JPhyloRef has been published in The Journal of Open Source Software, which should be cited when this tool is reused.

Phyloreferences are clade definitions in the form of well-structured machine-readable data with machine-interpretable semantics. Clade definitions define groups of organisms consisting of an ancestor and all of its descendants ("clades") based on shared ancestry. In evolutionary biology, clades are a fundamental unit for understanding evolution and describing biodiversity (see de Queiroz, 2007). Clade definitions therefore provide the theoretical foundation for the semantics of taxon concepts to be defined and reproducibly resolved within a hypothesis of evolutionary relationships, i.e., a phylogeny. The aim of Phyloreferencing is to structure and represent the semantics of a clade definition and phylogenetic hypotheses using ontologies and formal logic expressions (in the Web Ontology Language) such that machines can unambigiuously and reproducibly retrieve the nodes in a tree that match the semantics of the clade definition (if any). For more information on the motivation for phyloreferencing see Cellinese et al., preprint, and for more information on how phyloreferences are implemented, see the JOSS manuscript included in this repository, or phyx.js, a JavaScript library for creating phyloreferences using OWL ontologies.

JPhyloRef has two main goals:

1. The primary one is to facilitate automated testing that the semantics of the logical definitions imply ("resolve to") the correct nodes in the reference tree as clade ancestors. This is key in supporting quality control for the digitization of clade definitions from natural language text to a structured machine-interpretable representation. It also verifies that one of the theoretical foundational premises of phyloreferences, computational reproducibility, holds in practice.
2. The secondary goal is to enable integration with external tools that need to obtain the clade ancestor node(s) resulting from a given ontology of phyloreferences and reference tree(s). When run as part of an automated testing workflow, JPhyloRef reports test results in the cross-platform Test Anything Protocol format. When used to find clade ancestor nodes implied by logical clade definitions, results are returned as a JSON object. JPhyloRef uses the OWL API reference library for reading Web Ontology Language (OWL) ontologies, and for the actual ontology reasoning step it uses an external and configurable OWL reasoner.

Usage

JPhyloRef wraps the ELK reasoner and provides three ways in which it can be used to resolve phyloreferences:

• java -jar jphyloref.jar resolve input.owl: Resolves phyloreferences in input.owl and returns the nodes they resolve to in a JSON document.
• java -jar jphyloref.jar webserver: Starts a webserver that accepts ontologies for reasoning and provides the result as a JSON document.
  • Requests should be sent via POST to the /reason endpoint as an HTML form submission. The form can contain the ontology as a file upload in the jsonldFile element or as a string in the jsonld element. We will return a response in JSON with the results of the reasoning or with an error message.
  • You can also use the /version endpoint to test whether the software is working. It will report on the version of JPhyloRef, OWLAPI and reasoner being used.
  • Note that no content is served at /; you will need to use /version to test that the server is running.
• java -jar jphyloref.jar test input.owl: Test all the phyloreferences in input.owl by comparing their resolution with the expected resolution recorded in the file.

Detailed usage instructions are included in the JPhyloRef Usage document. Documentation of the source code is included as Javadoc comments, which are also available online at javadoc.io.

Command line options

Many command line options can be used for all included commands: - --jsonld or -j can be used to interpret the input file as a JSON-LD file rather than an RDF/XML file (resolve or test only). - --host [hostname] or -h can be used to set the hostname that the webserver should listen on (webserver only). - --port [port number] or -p can be used to set the port that the webserver should listen on (webserver only). - --reasoner [name] can be used to set the reasoner to use. The following reasoners are supported: - Elk 0.4.3 (elk) is an OWL 2 EL reasoner. Other reasoners for the OWL-EL profile may work but have not been tested. OWL-DL reasoners have been found to have insufficient performance.

Community guidelines

Development of JPhyloRef takes place in our GitHub repository. This includes an issue tracker for reporting any bugs you find or requesting any features you need. We welcome any pull requests to add additional features, tests or documentation. All new pull requests are tested with a continuous testing workflow.

Build and execution instructions

You will need Java and Apache Maven to build the software from source. We recommend installing these tools using a package manager, such as Homebrew on macOS. You can use jEnv instead if you want to install multiple Java versions on the same computer. Installation should set up the JAVA_HOME environment variable; if not, you will need to set it to point at the directory containing your Java installation.

Once you have downloaded the source code to your computer, you can compile and test the code by running mvn test.

JPhyloRef can be built from source by running mvn package from the root directory of this repository. This will create a JAR file in the target/ directory, which can be executed by running, for example:

$ java -jar target/jphyloref-1.2.0-SNAPSHOT.jar test src/test/resources/phylorefs/dummy1.owl

(Note that 1.2.0 should be replaced by the correct version number -- look for the Building JPhyloRef N.M.K-SNAPSHOT line in the output from mvn package, where N.M.K is the version, such as 1.2.0)

You can also download any published version of this software directly from Maven at https://search.maven.org/artifact/org.phyloref/jphyloref.

If you have Coursier installed, you can download and run JPhyloRef in one step by running:

$ coursier launch org.phyloref:jphyloref:1.1.1 -- test input.owl

Hosting a server with Webhook

JPhyloRef can be set up on a SLURM cluster using Webhook, allowing jobs to be executed on a separate computer from the web server.

Publishing to Sonatype OSSRH

To publish this package to the Sonatype OSSRH, we follow the workflow detailed on the Sonatype website. Note that you will need to set up a Maven settings.xml file with your GPG settings in order to sign the package for publication.

Once you're set up, you can run mvn clean deploy to publish the package to the OSSRH. If your version number ends in -SNAPSHOT, this will be published to the OSSRH Snapshots repository.