STEFTR: A Hybrid Versatile Method for State Estimation and Feature Extraction From the Trajectory of Animal Behavior.

Animal behavior is the final and integrated output of brain activity. Thus, recording and analyzing behavior is critical to understand the underlying brain function. While recording animal behavior has become easier than ever with the development of compact and inexpensive devices, detailed behavioral data analysis requires sufficient prior knowledge and/or high content data such as video images of animal postures, which makes it difficult for most of the animal behavioral data to be efficiently analyzed. Here, we report a versatile method using a hybrid supervised/unsupervised machine learning approach for behavioral state estimation and feature extraction (STEFTR) only from low-content animal trajectory data. To demonstrate the effectiveness of the proposed method, we analyzed trajectory data of worms, fruit flies, rats, and bats in the laboratories, and penguins and flying seabirds in the wild, which were recorded with various methods and span a wide range of spatiotemporal scales-from mm to 1,000 km in space and from sub-seconds to days in time. We successfully estimated several states during behavior and comprehensively extracted characteristic features from a behavioral state and/or a specific experimental condition. Physiological and genetic experiments in worms revealed that the extracted behavioral features reflected specific neural or gene activities. Thus, our method provides a versatile and unbiased way to extract behavioral features from simple trajectory data to understand brain function.

Measuring Spatiotemporal Dynamics of Odor Gradient for Small Animals by Gas Chromatography.

Odor is the most fundamental chemical stimulus that delivers information regarding food, mating partners, enemies, and danger in the surrounding environment. Research on odor response in animals is widespread, although studies on experimental systems in which the gradient of odor concentration is quantitatively measured has been quite limited. Here, we describe a method for measuring a gradient of odor concentration established by volatilization and diffusion in a relatively small enclosed space, which has been used widely in laboratories to analyze small model animals such as the nematode Caenorhabditis elegans and the fruit fly Drosophila melanogaster. We first vaporized known amounts of a liquid odorant 2-nonanone in a tank and subjected them to gas chromatographic analysis to obtain a calibration curve. Then, we aspirated a small amount of gas phase from a small hole on an agar plate and measured the odor concentration. By repeating this at different spatial and temporal points, we were able to detect a gradient of the odor concentration that increased over time. Furthermore, by applying these measured values to mathematical models of volatilization and diffusion, we were able to visualize an estimated dynamic change in odor concentration over an agar plate. Combining monitoring of odor concentration change in an agar plate with behavioral monitoring by machine vision will allow us to estimate how the brain computes information regarding odor concentration change in order to regulate behavior.

Calcium dynamics regulating the timing of decision-making in C. elegans.

Brains regulate behavioral responses with distinct timings. Here we investigate the cellular and molecular mechanisms underlying the timing of decision-making during olfactory navigation in Caenorhabditis elegans. We find that, based on subtle changes in odor concentrations, the animals appear to choose the appropriate migratory direction from multiple trials as a form of behavioral decision-making. Through optophysiological, mathematical and genetic analyses of neural activity under virtual odor gradients, we further find that odor concentration information is temporally integrated for a decision by a gradual increase in intracellular calcium concentration ([Ca2+]i), which occurs via L-type voltage-gated calcium channels in a pair of olfactory neurons. In contrast, for a reflex-like behavioral response, [Ca2+]i rapidly increases via multiple types of calcium channels in a pair of nociceptive neurons. Thus, the timing of neuronal responses is determined by cell type-dependent involvement of calcium channels, which may serve as a cellular basis for decision-making.

Modulation of different behavioral components by neuropeptide and dopamine signalings in non-associative odor learning of Caenorhabditis elegans.

An animal's behavior is modulated by learning; however, the behavioral component modulated by learning and the mechanisms of this modulation have not been fully understood. We show here that two types of neural signalings are required for the modulation of different behavioral components in non-associative odor learning in the nematode Caenorhabditis elegans. We have previously found that C. elegans avoid the repulsive odor 2-nonanone, and preexposure to the odor for 1h enhances the avoidance behavior as a type of non-associative learning. Systematic quantitative analyses of behavioral components revealed that the odor preexposure caused increases in average duration of straight migration ("runs") only when the animals were migrating away from the odor source within a certain range of bearing, which likely corresponds to odor decrement. Further, genetic analyses revealed that the genes for neuropeptide or dopamine signalings are both required for the enhanced odor avoidance. Neuropeptide signaling genes were required for the preexposure-dependent increase in run duration. In contrast, dopamine signaling genes were required not for the increase in run duration but likely for maintenance of run direction. Our results suggests that multiple behavioral components are regulated by different neuromodulators even in non-associative learning in C. elegans.