Multiple Domain Screen –
Comprehensive Lab Animal Monitoring System (CLAMSTM)
Material and Methods
(available through Columbus Instruments, USA, unless otherwise noted)
180C System Sample Pump
Oxygen and Carbon Dioxide Sensor Systems
Oxymax Equal Flow Multi-Animal Air Supply System
OPTO-M3 Activity Monitor
Slipper Tube Monitoring System
Control gas: 21% oxygen,
0.5% carbon dioxide, nitrogen to balance, BOC, Advantage Gases and Tools, USA
Magnesium Perchlorate, Sigma, USA
Calcium sulfate, anhydrate, Sigma, USA
Control Air Box
Animal housing units consisting of:
Plastic cage
Metal floor grid
Food hopper
Glass dryer column
IsoPad (fluid absorbent) sheets, Harlan, USA
Scale, Mettler-Toledo, USA
Steris CV Medicated Lotion Soap
Lab Diet, 6% fat, PMI Nutrition International, LLC, USA
Purified water, (Purification System D6311), Barnstead/Thermolyne, USA
Computer system Windows 2000
Communication port (included)
Analysis software
The comprehensive lab animal monitoring system (CLAMSTM) allows measurement of several physiological and behavioral parameters, including measurement of activity levels, volume of food and drink consumption, drink licking activity, temperature and metabolism. Since the animals are tested for 3 consecutive days it is possible to obtain data of responses to a novel environment and data of established diurnal patterns.
In preparation of the experiment, the components need to be checked and calibrated just prior to use: food hoppers and water bottles should be filled, cage floors lined with absorbent pads; water lines should have no kinks, connections to computer should be in place; test airflow into units, calibrate air composition and flow rate; initialize scales. Magnesium perchlorate is changed (wear gloves, work under hood) before every 5th experiment and the used material disposed of through hazardous waste disposal.
Animals are weighed prior to placing them into individual units between 11:30 a.m. and 1:00 p.m. on day 1, and at the end of the observation period. Mice that show negative results at 8-weeks of age are retested at one year of age to identify mutations that lead to moderate and/or late-onset obesity. At this time the mice are weighed again and their body-mass index is determined.
Algorithm development - Although simply examining data from each mouse for values that lie greater than 3 standard deviations above or below the control mean will detect potential mutants, this approach does not fully exploit the CCMS data set. We are currently developing a statistical algorithm for mutant detection that can incorporate multiple measures (Sen and Seburn). Briefly, the algorithm first "trains" on a data set from normal mice, then computes the Mahalanobis distance (Krzanowski, 1988) between subsequent individual mice and the training set. If this distance is "too" large the mouse is flagged as an outlier. The cut-off distance for flagging can be adjusted for a desired rate of false positives. The algorithm presently incorporates only 3 CCMS measures but will be expanded to incorporate additional parameters.
