Base Quality Score Correction Using Overlapping Read Pair Error Bins
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current base quality score correction methods, such as GATK-BQSR, fail to accurately distinguish between sequencing errors and sample-specific biological variations, leading to inaccuracies in sequencing data, particularly in high tumor mutation burden samples, and require significant computational resources.
Innovation Solution
A method that utilizes double-end sequencing data to differentiate between sequencing and non-sequencing errors by categorizing overlapping bases based on read direction, sequencing cycles, and dinucleotides, applying a locally weighted regression model to correct base quality scores.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If GATK-BQSR method is used to correct base quality scores, then sequencing errors can be corrected, but sample-specific biological variations and true sequencing errors cannot be distinguished, leading to inaccurate correction results
Solution Approach 1:
The patent segments the correction process by separating sequencing errors from biological variations through read pairing analysis. By examining overlapping bases between read pairs and comparing them against the reference genome, the method identifies and excludes sample-specific somatic mutations from correction, allowing accurate distinction between technical sequencing errors and biological truths.
Solution Approach 2:
The patent introduces an intermediary correction model that uses read pair information and overlapping base regions as mediators to filter out biological variations. This intermediary layer processes the raw quality scores through a model trained on read pair concordance, effectively mediating between the sequencer's quality assessments and biological reality.
2Measurement precision
If GATK-BQSR method is used for quality correction, then base quality scores can be adjusted, but computational time and resources increase significantly for high-depth and large-volume data
Solution Approach 1:
The patent performs preliminary actions by pre-processing read pairs to identify and flag potential biological variations before the main correction process. By pre-filtering reads based on read pair concordance and overlapping base analysis, the method reduces the computational burden during correction while maintaining accuracy.
Solution Approach 2:
The patent applies local quality correction by processing only the overlapping base regions between read pairs that show discordance with the reference genome. Instead of correcting entire genomes, the method focuses computational resources on localized error-prone regions identified through read pair analysis, significantly reducing processing time.
3Ease of manufacture
If all mismatches are treated as sequencing errors, then correction can be performed, but true biological variations and somatic mutations are incorrectly corrected
Solution Approach 1:
The patent introduces dynamic classification of mismatches by continuously evaluating read pair concordance and overlapping base regions during correction. The correction process dynamically adjusts its behavior based on whether a mismatch appears to be a sequencing error or biological variation, using read pair information to guide the correction decisions in real-time.
Data Source
AI summary
The current application discloses a method, apparatus, and electronic device for correcting base quality scores according to sequencing platform characteristics. This method extracts the overlapping bases, distinguishes sequencing errors from non-sequencing errors, and divides the extracted overlapping bases into different bins according to the read direction where the sequencing error base is located, the number of sequencing cycles, the dinucleotide in the sequencing direction, and the base quality score given by the sequencer. The sequencing error bases in each feature bin are counted, and the empirical quality value is calculated. A lowess model is used to perform polynomial fitting modeling on the RQS and EQS in the feature bins and to correct the original base quality score. The current application can distinguish between sequencing errors and non-sequencing errors, can more accurately reflect the sequencer preference, and can conduct modeling corrections, thereby comprehensively improving the credibility of base quality scores.