Cognition and Sensory perception
Introduction to the Issues
This overhaul was originally started by an e-mail to the GO neurobiology interest group. A User in Toronto was surprised that the most frequent gene products annotated to cognition were receptors such as olfactory receptors. On examining the annotations in MGI, I agreed that this was the case and that there may be problems in the structure of the graph in this area. Judy got me in touch with an expert at the lab and from her comments regarding cognition, perception and signaling, showed that the definition of perception was reasonably accurate, but there might be problems with its part_of children. She also pointed out that cognition is a controversial term to define in something like GO.
phone conference 5/7/2010
Tanya and I had a meeting with Daniele Merico on Friday May, 7 to discuss his issues with this area of the ontology and its annotations. The issues are:
- with the 2 children of cognition, 'sensory perception' and 'learning and memory'. The first is whether or not cognition is really the correct grouping term for these two processes. Perhaps we could get rid of cognition altogether and have these two processes as direct children of neurological process.
- should 'the detection of the stimulus' be a part_of sensory perception at all? Does sensory perception only include the 'higher order neural functions'? This is the way we (GO ontology developers) have always thought of it. If this is the case, then the olfactory receptors etc should not be involved in the perception per se.
- if we move the detection of the stimulus out from under sensory perception and merge it into its parent, then how do we convey the information that the olfactory receptors are involved in olfaction? Do we need more global terms for each of the senses?
- what should we do about learning and memory?
- mouse and human (and rat?) genes will need to be reannotated. At present, a large scale annotation was performed that associates all of the olfactory receptors with the sensory perception term. Based on Daniele's analysis, this is misleading. David will have to review the MGI annotation and get in touch with Emily and RGD once the ontology issues are resolved.
David and Tanya will schedule a second meeting with Elissa Chesler to discuss the issues with GO's current representation. We discussed this at an MGI GO meeting. Basically the representation of cognition is incorrect and needs to be reworked. The MGI phenotype group also attended this meeting. We will coordinate future efforts with them.
Third meeting David and Cindy Smith August 20,2010
After discussion it turned out that we did not need to perform that many rearrangements to GO to solve all of these issues. we moved sensory perception out from under cognition and then made sensory processing a child of cognition. It appears that the community has different definitions of cognition, but we thought that in the broadest sense, this move would work with everything Daniele and Elissa had said.