GPU-based real-time detection and analysis of biological targets using solid-state nanopores.

Area of Science:

• Nanotechnology and Biophysics
• Computational Biology and Data Analysis

Background:

• Nanoscale devices offer interfaces to biomolecules for studying biological interactions.
• Real-time computational analysis of measured data is crucial but often overlooked in these systems.

Purpose of the Study:

• To describe an adaptive approach for real-time recording and automated pattern analysis of single-molecule or single-cell events.
• To implement a computational system that mitigates overhead and accelerates data processing.

Main Methods:

• Utilized static and dynamic threshold techniques (baseline-tracker, moving average filtering) for automated data analysis.
• Implemented a real-time system with advanced I/O techniques to minimize execution stalls.
• Leveraged programmable graphics processing units (GPUs) for parallel pattern detection.

Main Results:

• The system enables real-time recording and automatic analysis of single-molecule or single-cell events.
• Dynamic threshold techniques provide noise suppression and precise pattern detection.
• The implemented system achieves five times faster data acquisition and pattern detection compared to the electrical measurement setup.

Conclusions:

• The developed adaptive system significantly enhances the speed and efficiency of analyzing nanoscale biological data.
• Real-time computational decision-making is a vital component for advancing single-event analysis in biological systems.
• Parallel processing on GPUs and optimized I/O are key to achieving high-speed data acquisition and pattern detection.