Supplementary data for: SarcAsM: AI-based multiscale analysis of sarcomere organization and contractility in cardiomyocytes

This repository contains the supplementary data for the publication "SarcAsM: AI-based multiscale analysis of sarcomere organization and contractility in cardiomyocytes". More details on samples see publication (https://www.biorxiv.org/content/10.1101/2025.04.29.650605v1). Training data set for sarcomere detection (training_data_sarcomeres.zip): Diverse set of fluorescent microscopy images with annotations for Z-bands, M-bands, sarcomere orientation map, sarcomere mask and cell mask. Data sources and details see Supplementary Table 2 of publication. Training data for ContractionNet (training_data_contractionnet.zip): Simulated and real data time series of contractions, including sarcomere and Z-band motion with ground truth contraction intervals. Test images (test_data.zip): Small test data set with diverse images and movies for testing SarcAsM. Validation data (validation_data.zip): Samples for validation of SarcAsM's ability to detect sarcomeres and measure sarcomere lenghts. For each image, we provide the manually annotated ground truth sarcomere mask and the manually measured sarcomere lengths (list and Fiji ROIs). Representative movies of drug treatment assays: data_chronic_drug_treatment.zip: 8-day live-movies (4 h frame interval) of ACTN2-citrine hiPSC-derived cardiomyocytes in 2D culture treated with Mavacamten (0.1µM, 1µM, 10µM, control) after day 3. data_acute_drug_treatment.zip: High-speed movies of ACTN2-citrine hiPSC-derived cardiomyocytes on micropatterned soft gel (20 kPa) treated with a small panel of cardiac drugs.