Check

To validate and, optionally, auto-fix and normalize BioPAX files or data (at some URL), submit a multipart/form-data HTTP POST request to this page

Parameters:

• file (actually, parameter name does not matter here, - simply submit an array of files) OR url (value: a URL to data in BioPAX format)
• retDesired - output format; values: "html" (default), "xml", or "owl" (modified BioPAX only, no error messages)
• autofix - false/true; try to fix some errors automatically (a new experimental feature; default is "false")
• normalize - false/true; return "normalized" BioPAX (a new experimental feature; default is "false")
• filter - set log level; values: "WARNING" (default, get both errors and warnings), "ERROR", and "IGNORE" (no problems at all ;))
• maxErrors - set the max. number of not fixed ERROR type cases to collect (some, but not all, warning and fixed cases will be also reported); value: a positive integer; "0" (default) means "unlimited", "1" - fail-fast mode, i.e., stop after the first serious issue, "10" - collect up to ten error cases, etc.

Output Formats:

• HTML - stand-alone HTML/JavaScript validation results page to save and view off-line
• XML - results follow the schema (XSD) There are different ways to convert the XML result to domain objects. JAXB works great (one can generate the classes from the schema; copy/modify the classes used by the Validator (sources here); or grab into your project the (latest snapshot) jar (from the BioPAX public repository) or simply add that repository and the org.biopax.validator:biopax-validator-core:2.0-SNAPSHOT:jar dependency to your pom.xml (though, this will automatically bring more dependencies to your project, e.g., paxtools-core, spring framework, etc...)
• OWL - modified BioPAX L3 data (fixed/normalized)

As an example, there is a basic BioPAX validator client module (it connects to the http://www.biopax.org/biopax-validator/check.html), and the sources are here (see test classes there as well:).