Related Experiment Video
Updated: Apr 21, 2026
10:50
Computational Modeling of Retinal Neurons for Visual Prosthesis Research - Fundamental Approaches
Published on: June 21, 2022
2.3K
Quantitative assessment of computational models for retinotopic map formation
J J Johannes Hjorth1, David C Sterratt2, Catherine S Cutts1
1Cambridge Computational Biology Institute, Department of Applied Mathematics and Theoretical Physics, University of Cambridge, Cambridge, CB3 0WA, United Kingdom.
Developmental Neurobiology
|November 5, 2014
Summary
Computational models of retinal axon guidance were assessed using a new framework and mouse data. Models combining molecular and activity cues require an anteroposterior gradient for accurate mapping in the superior colliculus.
Area of Science:
- Neuroscience
- Computational Biology
- Developmental Biology
Background:
- Retinal axon guidance to the superior colliculus (SC) involves molecular and activity-based cues.
- The precise interaction and contribution of these cues remain incompletely understood.
Purpose of the Study:
- To develop and utilize a quantitative framework for assessing computational models of retinocollicular mapping.
- To compare model performance against experimental data from the mouse retinocollicular system.
Main Methods:
- A computational framework was developed to evaluate existing models against curated experimental data.
- Four models combining Eph/ephrin gradients and activity-based mechanisms were tested.
- Models were assessed against five distinct genetic phenotypes, including wild type, knock-outs, and knock-ins.
Main Results:
- Two models accurately reproduced the Math5(-/-) anteromedial projection phenotype.
- Only one model could account for the Isl2-EphA3(ki/+) phenotype collapse point.
- A weak anteroposterior gradient in the SC was necessary to explain the residual order in the ephrin-A2,A3,A5 triple knock-out.
Conclusions:
- Testing retinotopic models against a comprehensive range of phenotypes is crucial for validation.
- Existing models require refinement, potentially incorporating additional guidance mechanisms, to fully explain observed retinocollicular map formation.
- A MATLAB software framework was developed to facilitate model assessment and simulation.
