ICONIP_with_acknowledgement.pdf (514.94 kB)
A biophysical model of the early olfactory system of honeybees
conference contribution
posted on 2023-06-09, 07:28 authored by Ho Ka Chan, Thomas NowotnyThomas NowotnyExperimental measurements often can only provide limited data from an animal’s sensory system. In addition, they exhibit large trial-to-trial and animal-to-animal variability. These limitations pose challenges to building mathematical models intended to make biologically relevant predictions. Here, we present a mathematical model of the early olfactory system of honeybees aiming to overcome these limitations. The model generates olfactory response patterns which conform to the statistics derived from experimental data for a variety of their properties. This allows considering the full dimensionality of the sensory input space as well as avoiding overfitting the underlying data sets. Several known biological mechanisms, including processes of chemical binding and activation of receptors, and spike generation and transmission in the antennal lobe network, are incorporated in the model at a minimal level. It can therefore be used to study how experimentally observed phenomena are shaped by these underlying biophysical processes. We verified that our model can replicate some key experimental findings that were not used when building it. Given appropriate data, our model can be generalized to the early olfactory systems of other insects. It hence provides a possible framework for future numerical and analytical studies of olfactory processing in insects.
History
Publication status
- Published
File Version
- Accepted version
Journal
International Conference on Neural Information ProcessingISSN
0302-9743Publisher
Springer VerlagExternal DOI
Volume
10637Page range
639-647Event name
ICONIP: International Conference on Neural Information Processing (2017)Event location
Guangzhou, ChinaEvent type
conferenceEvent date
14-18 November 2017ISBN
9783319700922Series
Lecture Notes in Computer ScienceDepartment affiliated with
- Informatics Publications
Full text available
- Yes
Peer reviewed?
- Yes
Legacy Posted Date
2018-02-20First Open Access (FOA) Date
2018-02-20First Compliant Deposit (FCD) Date
2018-02-20Usage metrics
Categories
No categories selectedLicence
Exports
RefWorks
BibTeX
Ref. manager
Endnote
DataCite
NLM
DC