Context-Aware Base Calling to Reduce Sequencing Errors
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing DNA sequencing technologies suffer from base calling errors that affect downstream bioinformatics analysis and research accuracy, necessitating a method to reduce the error rate and improve sequencing quality.
Innovation Solution
A method and apparatus for base calling that involves acquiring a first mapping relationship between correct/incorrect classification information of an initial base calling result and first sequencing information, using a first mapping relationship acquisition module and a correct/incorrect classification information determination module to determine the accuracy of base calling results based on contextual sequences across consecutive sequencing cycles.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional base calling methods are used, then the sequencing process is simple and fast, but the base calling error rate is high
Solution Approach 1:
The patent performs preliminary classification of sequencing cycles into correct and incorrect categories using trained classification models before final base calling. This preliminary action identifies potential errors early in the process, allowing for targeted correction without completely redesigning the base calling workflow, thus improving accuracy while controlling complexity
Solution Approach 2:
The patent introduces classification models as intermediary components between raw sequencing data and final base calling results. These models act as mediators that analyze sequencing information and provide corrected base calling results, improving accuracy without requiring fundamental changes to the underlying sequencing technology
2Measurement precision
If base calling error rate is reduced through multiple corrections, then sequencing accuracy improves, but processing time increases
Solution Approach 1:
Classification models are trained in advance on reference genome data, so that during actual base calling, only inference is needed rather than full re-training. This preliminary preparation enables fast real-time error detection and correction, improving precision without excessive processing time
Solution Approach 2:
The system uses the sequencing data itself and reference genome information to automatically identify and correct errors without requiring manual intervention or extensive external resources. The classification models self-adjust based on the input data patterns, enabling efficient processing
Data Source
Figure 1~3
Figure 4~5
AI summary
Disclosed are a method and an apparatus for base calling, a device, and a storage medium. The method includes: acquiring a first mapping relationship between correct/incorrect classification information of an initial base calling result of a sequencing cycle and first sequencing information of the initial base calling result of a sequencing cycle, where the first sequencing information includes first initial base calling information based on a designated sequencing cycle; and determining the correct/incorrect classification information of the initial base calling result of a sequencing cycle to be processed based on the first mapping relationship and the first sequencing information of the sequencing cycle to be processed. The method for base calling according to the method reduces the impact of a contextual sequence on the base calling and improves the accuracy of the base calling.