一种基于局部特征的单分子荧光事件识别方法
By employing a single-molecule fluorescence event recognition method based on local features, this method utilizes sliding window and deep learning techniques to extract single-molecule fluorescence trace fragment features. Combined with user-defined criteria, it solves the problems of inaccurate recognition results and reliance on human intervention in existing technologies, achieving rapid and accurate single-molecule event classification.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- TSINGHUA UNIVERSITY
- Filing Date
- 2024-07-25
- Publication Date
- 2026-07-17
AI Technical Summary
Existing methods for identifying single-molecule fluorescence events are labor-intensive and reliant on specialized knowledge when dealing with multi-state, rapid transitions, and complex fluorescence signals. Furthermore, deep learning models struggle to adapt to non-equilibrium conditions, resulting in low repeatability and large errors in the identification results.
A single-molecule fluorescence event recognition method based on local features is adopted. Local single-molecule fluorescence trace fragments are extracted by sliding window, and features are extracted by convolution and long-term capture units. Combined with user-defined classification criteria, stable and dynamic single-molecule events are identified.
It improves the accuracy and versatility of single-molecule fluorescence event recognition, enabling automatic analysis of high-throughput single-molecule FRET experimental data under equilibrium and non-equilibrium conditions, reducing human error, and achieving rapid and accurate single-molecule event classification.
Smart Images
Figure FT_1 
Figure FT_2 
Figure FT_3