Abdomen Tracking and 1-Minute Locomotion Activity Bins of Rhopalosiphum padi

Abdomen Tracking and 1-Minute Locomotion Activity Bins of Rhopalosiphum padi Description This repository hosts 96-hour video-tracking locomotion datasets for the bird cherry-oat aphid (Rhopalosiphum padi) maintained on an artificial diet under Light-Dark (16L:8D; LD) and constant darkness (DD). The deposit contains two primary components: Raw Tracking Coordinates (`.parquet` files in `.zip` archive): Frame by frame abdomen centroid tracks generated from the SLEAP pose estimation software. Tracking coordinates are spatially resolved, identity-assigned (aphids 1–6 per replicate chamber), and terminated at each individual's observed time of death or escape. Aggregated 1-Minute Activity Bins (`.csv`): Downstream locomotion dataset used in the activity analysis. Data was aggregated into 1-minute temporal bins, downsampled and filtered, and annotated time and treatment columns. Data in this Repository 16 individual master `.parquet` files (8 LD replicates and 8 DD replicates) along with a `README.md` detailing the data structure. `diet_1min_activity_downsample_rollingMean.csv`: Master CSV containing the processed 1-minute binned locomotion metrics across all replicates and treatments. Master Parquet Files Inventory The individual parquet files contained in the archive correspond to the following experimental replicates, photoperiod treatments, and host conditions: File Name Photoperiod Treatment Host Replicate agesync_7-29-25-rep1_master.parquet LD (16L:8D) Artificial Diet Rep 1 agesync_8-8-25-rep2_master.parquet LD (16L:8D) Artificial Diet Rep 2 agesync_8-19-25-rep3_master.parquet LD (16L:8D) Artificial Diet Rep 3 agesync_8-25-25-rep4_master.parquet LD (16L:8D) Artificial Diet Rep 4 agesync_9-11-25-rep5_master.parquet LD (16L:8D) Artificial Diet Rep 5 agesync_9-18-25-rep6_master.parquet LD (16L:8D) Artificial Diet Rep 6 agesync_1-2-26-rep7_master.parquet LD (16L:8D) Artificial Diet Rep 7 agesync_1-23-26-rep8_master.parquet LD (16L:8D) Artificial Diet Rep 8 agesync_10-3-25-rep1-DD_master.parquet DD (Constant Darkness) Artificial Diet Rep 1 agesync_10-23-25-rep2-DD_master.parquet DD (Constant Darkness) Artificial Diet Rep 2 agesync_11-17-25-rep3-DD_master.parquet DD (Constant Darkness) Artificial Diet Rep 3 agesync_11-28-25-rep4-DD_master.parquet DD (Constant Darkness) Artificial Diet Rep 4 agesync_12-16-25-rep5-DD_master.parquet DD (Constant Darkness) Artificial Diet Rep 5 agesync_12-26-25-rep6-DD_master.parquet DD (Constant Darkness) Artificial Diet Rep 6 agesync_2-11-26-rep7-DD_master.parquet DD (Constant Darkness) Artificial Diet Rep 7 agesync_2-18-26-rep8-DD_master.parquet DD (Constant Darkness) Artificial Diet Rep 8 Master Parquet Column Structure Each individual master parquet file contains the frame-by-frame tracked coordinate data with the following schema: Column Header Description global_frame Continuous frame index across the entire 96-hour assay. individual Assigned ID of the tracked aphid (1 through 6, left to right in the field of view) based on ROI spatial assignment. x Predicted horizontal coordinate (pixels) for the aphid abdomen. y Predicted vertical coordinate (pixels) for the aphid abdomen. score SLEAP pose estimation confidence score for the abdomen coordinate prediction. treatment Experimental lighting regimen (LD or DD). replicate Video replicate identifier corresponding to the experimental run. video_id Source recording chunk identifier from the acquisition pipeline. 1-Minute Activity Bins Structure This CSV file has the following columnar structure: Column Header Description treatment Photoperiod and host treatment (i.e., LD_diet) replicate Experimental replicate identifier (1-8) individual Chamber arena/individual identifier (1-6. left to right in the field of view). minute_bin Sequential 1-minute interval index across the recording (ending at the time of death for each individual). filtered_distance Total filtered trajectory distance in each minute bin (in pixels; 281 pixels per cm). tracked_frames Number of frames with tracking coordinates for the abdomen (as some frames don't have any tracks due to occlusions, low accuracy, etc.) tracking_coverage Fraction of tracked frames as compared to the total possible number of frames in each minute bin. is_active Boolean flag if activity was greater than the bout threshold; 1 = active, 0 = inactive. ZT_continuous Zeitgeber time (in hours) across the entire 96 hour assay. ZT_day Zeitgeber time (in hours) wrapped modulo 24 hours. Code and Reproducibility The code used to generate these datasets and the downstream analysis can be found at the GitHub repositories: H5 to Master Parquet Extraction: github.com/dankunk/Rpadi_Video_Tracking_Analysis/tree/master/H5toParquet Activity Binning & Filtering Pipeline: github.com/dankunk/Rpadi_Video_Tracking_Analysis/tree/master/GenerateActivityBins