Development

This page describes the key features of the current version of the converter and outlines the expected future developments and the corresponding milestones.

Current version: PD2AF 1.0

The implemented logic takes into account on the work by Vogt and coauthors (Vogt et al., 2013, doi: 10.1186/1752-0509-7-115) and further develops the logic to minimise the number of complexes in AF version and to introduce pattern recognition functionalities.

What is new in the Milestone 1 PD2AF version 1.0:

  1. The code is written in Lisp programming language and is accessible on GitHub.
  2. Added pattern recognition for removing “intermediate” complexes.
  3. Added pattern recognition for some types of inhibition shown in details without using the inhibition arc.

Development plan

The development is planned in a stepwise manner with the following milestones.

PD2AF Milestone #1 Basic - “Enhanced” version - Automatic

The first milestone version is designed to provide an accessible open-source code and an online service. Milestone 1 has minimal translation functionalities enhanced by the identification of active entities. This identification allows improving the conversation while taking into account the context - the connected events. In particular, cases of “hidden” inhibition can be identified and converted to the Activity Flow view.

After completing this step, building a collection of PD-AF pairs of bricks can be initiated. In the future, this collection of pairs will ensure stable conversion between the Process Description and Activity Flow languages in both directions.

PD2AF Milestone #2 Intermediate - “Adviser” version - Semi-automatic

For complex and extensive PD maps it will not always be possible to avoid keeping complexes in AF version. For Milestone 2 we anticipate the development of a version that will not be completely automatic but will offer options for possible conversion of difficult modules. The user would be able to choose the most suitable option from a set.

PD2AF Milestone #3 Advanced - “Informed guess” version - Automatic

Milestone 3 aims to allow the algorithm to make a decision for difficult modules while taking into account information from selected pathway and protein-protein interaction databases. This “informed guess” functionality will allow minimising manual check for conversion of extensive maps.