nanite

nanite analyzes atomic force microscopy (AFM) force-distance (FD) curve data to extract mechanical properties and perform quality-based automated sorting for quantitative AFM imaging and mechanical characterization.

Key Features:

• Data Import and Preprocessing: Supports loading of AFM force-distance (FD) curve data and initial preprocessing steps.
• Tip-Sample Separation and Baseline Correction: Performs automatic tip-sample separation and corrects baseline drifts in FD curves to remove artifacts from passive cell movement, adhesion, or inadequate tissue attachment.
• Contact Point Retrieval and Model Fitting: Identifies contact points in FD curves and fits mechanical models to quantify properties such as stiffness and Young's modulus.
• Automated Sorting Using Supervised Learning: Uses supervised learning to correlate subjective quality ratings with extracted features, achieving mean squared error below 1.0 rating points and classification accuracy above 87% for distinguishing good versus poor-quality curves.
• Quantitative Imaging Enhancement: Enables incorporation of data quality as an additional dimension in AFM-based quantitative imaging and FD maps.

Scientific Applications:

• High-throughput FD dataset analysis: Automated evaluation and filtering of large sets of FD curves for reproducible mechanical property estimation.
• Quantitative AFM imaging and FD maps: Integration of quality-filtered FD data into maps for spatially resolved Young's modulus measurements.
• Biological tissue mechanics: Demonstrated application to quantify Young's modulus of the zebrafish spinal cord under different classification thresholds.

Methodology:

Computational steps explicitly include data import, automated preprocessing comprising tip-sample separation, baseline correction and contact point retrieval, application of model fitting to processed curves, and supervised-learning-based quality assessment for automated sorting.