Using Neurosim
Laboratory
Neurosim can in part replace normal laboratory experiments. As well as the obvious advantages of reduced consumable costs and animal usage, Neurosim enables students to do virtual experiments that would simply not be feasible in a normal teaching laboratory. This aids the understanding of fundamental principles, and gives practice in interpreting and analysing realistic-looking neurophysiological data similar to that which they encounter when reading the research literature.
Parameters can be hidden so that students have to discover them by experiment. This gives them the opportunity to design and carry out their own experiments in a risk-free environment - they can start again with zero cost (except their time) if their ideas do not pan out as they hoped.
Student Projects
Neurosim is well suited for use in student project work. In particular, the Advanced HH and Network modules can be used to develop many of the types of circuit described in the computational neuroscience literature, including coordinated locomotory circuits, pattern recognition circuits, memory-related networks and many others.
Lectures
Neurosim can be used as a visual aid during lectures to present dynamical events that cannot be adequately illustrated with conventional techniques.
Tutorials, and Tests
Neurosim can generate simulated neurophysiological data for off-line use in tutorial discussion, or for problem-solving questions for exams. The parameter-hiding facility means that it can also be used for on-line tests when students have access to Neurosim within the test environment.
Remote Teaching
Neurosim can be delivered to students from a central university server using a VDI-like program such as Aporto or AppsAnywhere. This enables students to run the program as if it were installed on their own computer (either a PC or Mac).
Alternatively, a tutor can run Neurosim on their own PC (or Mac using Bootcamp), and then invite tutees to join a conference call using Microsoft Teams or Zoom or an equivalent program. The tutor can share their screen, and then give individual tutees control, so that a student can carry out experiments using Neurosim as though it were running on their own computer. The student can be running the conference software on a Mac computer, even though Neurosim is a PC program. A short video illustrating a try-out of this usage can be seen here.
Graduate Catch-Up
The supplied tutorial exercises form a good "getting up to speed" program for graduate students or post-docs who are moving into an electrophysiological research program from a different area of neuroscience.
Advice to Tutors
- Neurosim is not a stand-alone self-teaching system for undergraduates. The supplied tutorial exercises explain some of the more advanced topics, but it is assumed that students have had some instruction in basic theory before they undertake the simulations. It is good practice in most cases to have demonstrators and tutors on hand, at least initially, to discuss the results with students.
- In a laboratory setting it is often advantageous to have the students working in pairs or small groups. Within-group discussion and idea-sharing can definitely enhance the experience for the students (and alleviate some of the pressure on instructors!).
- Neurosim obviously cannot teach the technical and practical skills involved in performing experiments. For this there is no substitute for actually doing real experiments.
- A more subtle issue with Neurosim is that its experiments always work, and always yield interpretable results (barring program bugs!). This is excellent for teaching and understanding, but it can give students an over-sanguine and under-critical idea of how real science works. For this reason, I think it is important to continue to expose all students to some real laboratory work, where they are likely to encounter occasional unexpected and uninterpretable outcomes.
- Finally, students should be reminded that just because a computer model produces realistic-looking output, it does not prove that the real system actually uses the same mechanism as the model. The point was rather well made by Hodgkin and Huxley themselves in their classic 1952 paper:
This is an important point for all computer models, not just Neurosim....the success of the equations is no evidence in favour of the mechanism of permeability change that we tentatively had in mind when formulating them”