Anchor-Based Sequence Alignment for NGS Variant Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional next-generation sequencing (NGS) alignment methods often result in misalignments and soft-clipping, particularly when variants are near the end of sequencing reads, leading to missed detection of important sequence variants such as insertions and deletions.
Innovation Solution
The method involves performing a first alignment of NGS sequencing reads to a reference genome, followed by the addition of anchor sequences to both sides of the reads and the target sequences. A second alignment is then performed, which forces the reads to align correctly by creating a high score for the anchor sequences, thereby preventing misalignments and soft-clipping.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If conventional alignment methods perform soft-clipping when detecting poor alignment, then alignment speed is improved, but variant detection precision deteriorates because sequences following the poor alignment position are hidden from analysis
Solution Approach 1:
The alignment process is divided into two distinct stages: first alignment for initial mapping and speed, followed by second alignment for precision variant detection. This segmentation allows each stage to optimize for its specific purpose without compromise.
Solution Approach 2:
The first alignment serves as a preliminary action that quickly identifies potential variant regions. This preliminary mapping enables the second alignment to focus computational resources on regions of interest, improving both speed and precision.
2Measurement precision
If anchor sequences are added to extended sequencing reads and target sequences for second alignment, then variant detection precision is improved, but device complexity increases
Solution Approach 1:
Anchor sequences serve as intermediary elements that mediate between the sequencing read and the target sequence. These anchors provide stable reference points that facilitate precise alignment without requiring complex alignment algorithms throughout the entire sequence.
Solution Approach 2:
Instead of applying complex alignment procedures uniformly across the entire sequence, the invention applies high-precision alignment only to local regions containing anchor sequences. This localized approach maintains precision while reducing overall computational complexity.
3Reliability
If the length of anchor sequences is increased to improve alignment stability, then alignment reliability is improved, but processing time increases
Solution Approach 1:
The invention uses partial anchor sequences rather than requiring complete sequence alignment for reliability. By using sufficiently long anchors (at least 4 nucleotides) without extending to the full read length, the system achieves reliable alignment while avoiding excessive processing time.
Data Source
Figure 1(a)~1(c)
AI summary
A method to align a next generation sequencing read to a reference sequence includes: (a) receiving a sequencing read; (b) performing a first alignment of the sequencing read to a reference sequence so as to identify a target sequence within the reference sequence whereto the sequencing read maps; (c) selecting a first and a second anchor sequence; (d) attaching the first anchor sequence to the upstream region of the sequencing read and the second anchor sequence to the downstream region of the sequencing read so as to generate an extended sequencing read; (e) attaching the first anchor sequence to the upstream region of the target sequence and the second anchor sequence to the downstream region of the target sequence, so as to generate an extended target sequence; (f) performing a second alignment of the extended sequencing read to the extended target sequence, so that the second alignment is more correctly mapped to the target sequence than the first alignment; (g) identifying a position where one or more bases between the extended sequencing read and the extended target sequence are different based on the second alignment result; and (h) calling a variant based on the identification in step (g).