ABSTRACTWe describe a novel quantitative home cage monitoring (HCM) approach for dissecting spontaneous patterns of ingestive and locomotor behaviors into a hierarchically organized series of behavioral facets or endophenotypes. Fine-grained analyses of a large multimodal 16-strain behavioral dataset collected from 169 mice revealed bouts of feeding, drinking and locomotor behaviors occurring within animals’ Active States. We have automated the detection of these bouts and their discrete properties including bout sizes, rates, durations, and intensities. We have developed a hierarchically organized model of behavioral organization enabling analysis of relationships among Active/Inactive State properties and those of feeding, drinking and locomotor bouts. Robust and analogous patterns of interrelationships among these endophenotypes were found for feeding, drinking behaviors, and these differed markedly from those for locomotor behaviors. For feeding and drinking, patterns of reciprocal relationships were observed for pairs of endophenotypes at multiple hierarchical levels. Moreover, endophenotype variability was highest at lowest hierarchical levels progressively diminished at higher levels, so that variability of gross levels of food and water intake were much less than those of their lower level determinants. By contrast, interrelationships among locomotor endophenotypes differed markedly from those of ingestive behavior. Altogether, these findings raise the possibility that behavioral regulation of food and water intake may make an important contribution to the homeostatic maintenance of energy and volume balance.
Real-Time Compression for Tactile Internet Data Streams
