About
Paxtools is a library for accessing and manipulating BioPAX. Software tools that use BioPAX, such as exporters, importers, analysis algorithms or editors can use Paxtools as their core BioPAX API. Paxtools supports BioPAX Level 2, and can automatically convert Level 1 models to Level 2. The latest release candidate for BioPAX Level 3 is fully supported, but it is still in beta as BioPAX Level 3 is not finalized. Using Paxtools, a user can read a BioPAX model from a file or create a new model in memory. A model is a container for BioPAX elements, and has methods for querying and retrieving its contents. In a BioPAX model, elements are represented as plain java beans, with getter and setter methods for their BioPAX properties. A user can create and add new BioPAX elements to the model, remove existing elements or modify their properties. It is also possible for users to traverse the corresponding BioPAX graph to retrieve interesting subgraphs.
Paxtools provides a domain object model (DOM) with strong typing for domains and ranges of the properties. It checks whether cardinality constraints are observed. Rules that were documented but could not be represented formally in BioPAX OWL are also validated against. Paxtools fails fast if these rules are violated. Additionally, Paxtools checks for documented best practices and logs potentially problemmatic usages.
Paxtools comes with a state-of-the-art persistence and searching layer using Java Persistence API (JPA) and Lucene. This enables storing and querying particularly large models and concurrent editing of the same model by multiple users. Lucene allows efficient fulltext querying.
Paxtools has an experimental support for integrating two pathways from different sources by identifying interactions that are similar. It is often difficult to make an exact matching between entities in several cases- for example the same state of the proteincan be represented as the active state in one pathway database and doubly phosphorylated in another one. An iterative two-step process fuzzy-matching entities followed by finding interactions that have similar participants however provides quite reasonable pathway alignments. This facility can be used for cross-validating pathways from different sources or as a first step for integrating overlapping pathways. Paxtools has also several other useful functionalities such as reflection based access, traversal methods, and programmatic access to Pathway Commons web service API.
Documentation
User's Guide
Paxtools User's Guide can be downloaded in PDF format from the following link:Download user's guide
Please note that, the latest version of this file can always be obtained from the CVS repository, under the doc directory:
http://biopax.cvs.sourceforge.net/biopax/Paxtools/
Download & Development
The Paxtools core is developed at Computational Biology Center at MSKCC, and the persistency layer is developed by INOH database.
Development of Paxtools is maintained under SourceForge.net. Information regarding development and links for downloading can be found under the SourceForge project web site:
Releases
cbio.mskcc.org