KaicoTracker: a robust and automated locomotory analysis for baculovirus-infected silkworm larvae

Abstract

AbstractBombyx mori nucleopolyhedrovirus (BmNPV) is a pathogen of the domestic silkworm, Bombyx mori, whose infection drastically alters larval locomotory patterns and activity. As B. mori is the only insect that has been completely domesticated, its larvae are typically inactive. In contrast, BmNPV-infected larvae exhibit vigorous movement. Recently, we developed a method for locomotory analysis of BmNPV-infected B. mori larvae, which yielded high-resolution data on BmNPV-induced host locomotory alteration. However, this method had a disadvantage when analyzing the behavior of uninfected larvae, as they were typically inactive and difficult to distinguish from the background. In addition, the analysis processes involve many manual steps, which reduces analysis throughput. We have developed KaicoTracker, a new tool for automated locomotory analysis of BmNPV-infected B. mori larvae, using Python and OpenCV. This tool is an all-inclusive package that tracks and quantifies larval locomotion and verifies the accuracy of the tracking. Larval locomotion is tracked by motion tracking function based on k-Nearest Neighbor implemented in OpenCV. Based on the tracking results, the positions of the larvae are determined, and locomotory distance, speed, duration, and frequency are calculated. The larval positions are marked on the original video, allowing tracking precision to be examined. By analyzing previously obtained data, we confirmed that this tool provides a robust analysis of inactive larvae with fewer manual steps.