Phase-State Model for Sequencing-by-Synthesis Signal Correction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Sequencing-by-synthesis technologies face challenges in accurately estimating signal correction parameters, particularly phasing effects and signal droop, which hinder the ability to make precise base calls due to noise introduction and phase synchrony loss among template strands.
Innovation Solution
A phase-state model is constructed to simulate the sequencing process, incorporating signal correction parameters such as incomplete extension rate, carry forward rate, and signal droop, which adjusts predictions based on comparisons with measured signal data to improve fit and accuracy of base calling.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional base calling algorithms are used without accurate signal correction parameters, then the base calling process is simpler and faster, but the accuracy of base calls deteriorates due to uncorrected phasing effects and signal droop
Solution Approach 1:
The patent applies preliminary action by pre-calculating and storing signal correction parameters (phasing parameters and signal droop parameters) before the base calling process. These parameters are determined from training data and stored in a lookup table, allowing the base calling algorithm to quickly retrieve and apply corrections without performing complex real-time calculations, thus improving accuracy while maintaining computational efficiency
Solution Approach 2:
The patent introduces an intermediary component - the signal correction parameters (phasing parameters and signal droop parameters) - that mediate between the raw sequencing signals and the base calling process. These parameters act as a bridge that translates complex signal degradation patterns into correctable forms, allowing the base calling algorithm to achieve high accuracy without directly modeling the complex physical processes of phasing and signal decay
2Measurement precision
If signal correction parameters are estimated from limited training data, then the model training is faster and requires less computational resources, but the accuracy of parameter estimates deteriorates leading to increased noise in signal analysis
Solution Approach 1:
The patent applies preliminary action by performing extensive model training and parameter estimation before actual sequencing runs. Training data from multiple sequencing runs is collected and processed in advance to determine accurate phasing parameters and signal droop parameters. These pre-computed parameters are stored and reused for subsequent base calling operations, allowing accurate parameter estimation without incurring the computational cost during time-critical sequencing operations
Solution Approach 2:
The patent applies partial or excessive action by using a large volume of training data from multiple sequencing runs to estimate signal correction parameters. By aggregating data from many runs, the model over-determines the parameters, providing robust estimates that are more accurate than what could be obtained from a single run. This excessive use of training data ensures parameter accuracy while the pre-computation approach maintains efficiency during actual sequencing
Data Source
Figure 1
Figure 2
Figure 3A
AI summary
A method of obtaining a more accurate estimate of a signal correction parameter(s) in sequencing-by-synthesis operations, such as incomplete extension rates, carry forward rates, and/or signal droop rates. The sequencing operation produces signal data. A model is constructed to simulate a population of template strands as it undergoes the sequencing process and becomes divided into different phase-states as the sequencing-by-synthesis progresses. For example, the model may be a phase-state model. The output from the model is used to adjust the signal correction parameter(s). For example, the model may be fitted to the signal data. This fitting results in a more accurate estimate of the signal correction parameter(s). In another embodiment, the signal droop rate is modeled as a decaying function and this decaying function is fitted to the signal data to obtain an improved estimate of the signal droop rate.