TUTORIAL

SEE: a software-supported Strategy for the Exploration of Exploration

Dina Lipkind, Ilan Golani and Yoav Benjamini (Tel Aviv University, Tel Aviv, Israel)

Benefits
You will learn about new software-based techniques for the analysis of exploratory behavior of laboratory animals, relevant for basic research as well as behavioral phenotyping studies:

  • SEE analysis provides a large number of behavioral measures (endpoints) that were shown to be discriminative between mouse strains and replicable across laboratories.
  • Endpoints are calculated automatically from the output of tracking systems like Noldus EthoVision.
  • Endpoints are based on natural behavior patterns that are defined algorithmically without the need for a human observer, and, consequently, without subjective bias.
  • SEE analysis is high throughput, allowing fast analysis of large quantities of data. One half hour Open Field session, providing several dozens of endpoints, can be analyzed in several minutes.
  • A publicly accessible database of EthoVision output files from Open Field sessions of many commonly used and genetically diverse inbred mouse strains allows researchers to compare their results to results from other laboratories.
  • SEE provides statistical tools that address current key problems in behavioral phenotyping.

Features
SEE consists of several stand-alone user-friendly, publicly available programs written in C++ and Mathematica routines. It includes:

  • SEE Path Smoother (SPSM). This software provides a robust smoothing algorithm that reduces tracking noise. It combines the use of a running median smoother that captures even the shortest stops, and a subsequent smoothing method called Lowess, which assures a robust smoothing.
  • SEE Path Segmentor (SPSG). This software divides the behavior into stops and progression segments. The input for this software is the output of SPSM.
  • SEE Endpoint Manager (SEM) calculates several dozens of ethologically relevant endpoints per data file.
  • SEE package. This is a Mathematica-based software package that allows the visualization of the data, the construction of hypotheses concerning the organization of behavior, and the development of new behavioral endpoints.
In the tutorial we will demonstrate and practice the use of SEE, which is straightforward and easy to learn.

Audience
Users of video tracking systems in behavioral research.

Presentation
Lecture and practical demonstration followed by questions and answers session.

Instructor resume
The instructors belong to a research group that has developed SEE, and used it on rats and on 10 of the most commonly used inbred mouse strains in three laboratories in Israel (Tel Aviv University) and the USA (National Institute of Drug Abuse and Maryland Psychiatric Research Center, Baltimore). Prof. Ilan Golani and Dr. Dina Lipkind are both zoologists, Dr. Yoav Benjamini is a statistician. More information about their research group is available at www.tau.ac.il/~ilan99.

Sign up
To sign up for this tutorial, please send an email to Ilan Golani. Feel free to ask questions in advance, thus allowing the instructor to tailor the content of the tutorial to the interest of the participants.


Last updated: 5 July 2002