DNA Read Clustering Using Edit Distance and Hash Bucketing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Polynucleotide sequencers produce error-prone DNA reads, making it challenging to accurately cluster sequences due to the high volume and noise in the data, which complicates error correction and sequence inference.
Innovation Solution
The method involves clustering DNA reads based on edit distance and hash values, using iterative processes to group similar sequences together, with the aid of binary signatures and randomized locality-sensitive hashing, to efficiently organize reads representing the same original DNA strand.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If DNA reads are clustered based on edit distance to group sequences from the same original strand, then clustering accuracy is improved, but computational expense increases due to the high volume and noise in the data
Solution Approach 1:
The patent segments the computational task by first grouping DNA reads into buckets based on their prefixes (first k nucleotides), then performing edit distance calculations only within each bucket. This segmentation reduces the number of pairwise comparisons needed, as reads from different buckets are unlikely to be similar. The iterative process further segments the problem by re-bucketing and re-clustering until convergence, making the overall computation more efficient while maintaining accuracy.
Solution Approach 2:
The patent performs preliminary grouping of DNA reads into buckets based on prefix matching before conducting the computationally intensive edit distance calculations. This preliminary action filters out obviously dissimilar reads that would not belong to the same cluster, reducing the search space for subsequent clustering iterations. The prefix-based bucketing is computed once and reused across multiple iterations, saving computational resources.
2Manufacturing precision
If iterative clustering processes are used to improve sequence accuracy, then error correction quality is improved, but processing time increases
Solution Approach 1:
The patent employs periodic action through iterative clustering processes that repeat a cycle of: (1) computing edit distances between reads in the same bucket, (2) forming or merging clusters based on distance thresholds, (3) removing clustered reads from further consideration, and (4) re-bucketing remaining reads. This periodic iteration continues until all reads are clustered or no further improvements are possible, systematically improving accuracy while avoiding unnecessary computations by stopping when convergence is reached.
3Productivity
If hash values and binary signatures are used to approximate edit distance, then computational efficiency is improved, but measurement precision may be reduced
Solution Approach 1:
The patent uses hash values and binary signatures as intermediary representations of DNA reads to approximate edit distance calculations. Instead of computing the full edit distance between all read pairs, the system computes hash values from binary signatures (which encode the presence/absence of k-grams) and uses the Hamming distance between these signatures as a proxy. This intermediary approach provides a computationally efficient approximation that guides the clustering process, with more precise edit distance calculations performed only when needed within buckets.
Data Source
AI summary
A technique for clustering DNA reads from polynucleotide sequencing is described. DNA reads with a level of difference that is likely caused by errors in sequencing are grouped together in the same cluster. DNA reads that represent reads of different DNA molecules are placed in different clusters. The clusters are based on edit distance, which is the number of changes necessary to convert a given DNA read into another. The process of forming clusters may be performed iteratively and may use other types of distance that serve as an approximation for edit distance. Well clustered DNA reads provide a starting point for further analysis.