AF development

Contents

AF nodes

Since the release of SBGN Activity Flow Level 1 specification, we have received some feedback, particularly about the biological activity glyph. Many of the concerns are reasonable (see details below).

Last April, we conducted a poll on the topic. The SBGN AF Level 1 specification was drafted based on the outcome of the poll. Unfortunately, we had very few people voted in the poll. Many of the feedback came after the poll was closed and even after the specification was released. We would like to re-open the discussion in order to ensure that the specification will be supported by the community. This is not a poll, but rather a discussion. The proposals are described in this file. Please review and provide your feedback. If you would like to propose a new representation, please feel free to send out to the discussion list. We hope to conclude the discussion sometime in March. If we have a consensus, that will be great. If the outcome is not clear, we will conduct a poll then.

Here we go.

One consensus we have reached is that AF is about influences among the activities, therefore the nodes in AF diagram should represent activities rather than entities. The problem arises when one wants to decorate the activity node to display the type of entity that the activity comes from.

The goal is to create biological activity nodes that are

  • unambiguous
  • ontologically accurate
  • consistent with representations in all three SBGN languages

Current specification

  • The activity node is represented by a rectangle.
  • Entity decoration uses unit of information to indicate the type of entity where the activity comes.
  • Examples: EGFR; PPAR; TGF-beta
  • Advantages: It is ontologically correct to use a unique glyph to represent biological activity, which is an event.
  • Disadvantages: The unit of information decoration can be too small to differentiates, e.g., a macromolecule vs a nucleic acid feature or a complex. It can be useless when the node itself is small in a very large map.

Proposal 1

  • The biological activity node is represented by a rectangle.
  • The entity decoration is drawn the same way as in PD. The AN is placed on the line of the entity glyph to indicate that the activity is derived from the entity.
  • All modulation arcs (positive influence, negative influence) must be connected to the activity nodes, but not the entity glyphs. Therefore, the entity glyph is served more like the container.
  • Examples: EGFR; PPAR; TGF-beta
  • One problem with this proposal is that the activity node looks quite similar to the unit of information in the SBGN PD.

Proposal 2

It is similar to proposal 1, except that the biological activity nodes are placed inside rather than on the edge of the entity glyph.

Proposal 3

  • The biological activity node is represented by a rectangle when the entity type of the activity is not illustrated. When the entity type needs to be illustrated, glyphs same to the EPN in PD are used to indicate the type of entity the activity is from. However, here, these nodes represent biological activity but not entity per se.
  • Examples: EGFR; PPAR; TGF-beta
  • The major criticism about this proposal is that it is ontologically ambiguous.

Please send all comments to SBGN discussion list

Retrieved from "http://sbgn.org/AF_development"

This page was last modified 17:32, 2 March 2010.

Please use our issue tracker to send us suggestions and problem reports about this website.
This page was last modified 17:32, 2 March 2010.