PT - JOURNAL ARTICLE AU - Shuhei J. Yamazaki AU - Yosuke Ikejiri AU - Fumie Hiramatsu AU - Kosuke Fujita AU - Yuki Tanimoto AU - Akiko Yamazoe-Umemoto AU - Yasufumi Yamada AU - Koichi Hashimoto AU - Shizuko Hiryu AU - Takuya Maekawa AU - Koutarou D. Kimura TI - Experience-dependent modulation of behavioral features in sensory navigation of nematodes and bats revealed by machine learning AID - 10.1101/198879 DP - 2017 Jan 01 TA - bioRxiv PG - 198879 4099 - http://biorxiv.org/content/early/2017/10/12/198879.short 4100 - http://biorxiv.org/content/early/2017/10/12/198879.full AB - Animal behavior is the integrated output of multiple brain functions. However, understanding how multiple brain functions affect behavior has been difficult. In order to decipher dynamic brain functions from time-series of behavioral data, we developed a machine learning strategy that extracts distinguishing behavioral features of sensory navigation. We first investigated experience-dependent enhancement of odor avoidance behavior of the nematode Caenorhabditis elegans. We segmented worms' trajectories during olfactory navigation into two behavioral states, analyzed 92 features of the states, and automatically extracted 9 distinguishing features modulated by prior odor experience using a statistical index, the gain ratio. The extracted features included ones previously unidentified, one of which indicated that the prior odor experience lowers worms' behavioral responses to a small increase in odor concentration, causing enhanced odor avoidance. In fact, calcium imaging analysis revealed that the response of ASH nociceptive neurons to a small odor increase was significantly reduced after prior odor experience. In addition, based on extracted features, multiple mutant strains were categorized into several groups that are related to physiological functions of the mutated genes, suggesting a possible estimation of unknown gene function by behavioral features. Furthermore, we also extracted behavioral features modulated by experience in acoustic navigation of bats. Thus, our results demonstrate that, regardless of animal species, sensory modality, and spatio-temporal scale, behavioral features during navigation can be extracted by machine learning analysis, which may lead to the understanding of information processing in the brain.SIGNIFICANCE STATEMENT Behavior is the most important output of brain activity, and its recording has become easy because of the development of small and inexpensive cameras and small GPS devices. However, these "behavioral big data" have been used to calculate very simple indices, such as speed, direction, and goal arrival rate. In this study, we analyzed animal behavior using machine learning (also known as "artificial intelligence") and found specific behavioral features related to navigation in worms and bats. We also found activity changes in nerve cells that were reflected in the worm's behavioral changes. Thus, our results demonstrate that artificial intelligence can be used to find characteristics of animal behavior that would eventually help us understand how the brain works.