Noisy DNA Read Clustering with Hash-Based Edit Distance Screening
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Polynucleotide sequencers produce errors in DNA sequencing, making it difficult to accurately identify and correct errors in the DNA sequence data, especially due to the high volume of data generated and the random distribution of errors introduced during amplification and sequencing processes.
Innovation Solution
The method involves clustering DNA reads based on edit distance and hash values to group reads that represent the same original DNA strand, using iterative processes to reduce computational expense and improve accuracy, by designating representative reads and approximating edit distance with Hamming distance calculations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If DNA reads are clustered based on edit distance to group reads from the same original strand, then clustering accuracy is improved, but computational complexity and processing time increase significantly
Solution Approach 1:
The patent introduces hash values as an intermediary representation of DNA reads. Instead of directly comparing all pairs of reads using computationally expensive edit distance calculations, the system first computes hash values for each read and clusters reads with matching or similar hash values. This intermediary hashing step dramatically reduces the number of edit distance computations needed while maintaining clustering accuracy, as hash values serve as efficient proxies for read similarity.
2Reliability
If all pairs of DNA reads are compared using edit distance, then clustering completeness is improved, but processing time becomes prohibitive for large datasets
Solution Approach 1:
The patent segments the clustering process into multiple stages: first, reads are grouped by hash value matches; second, representative reads from each hash group are selected; third, edit distance is computed only between representative reads from different hash groups. This segmentation transforms an O(n²) problem into a much more efficient process by dividing the dataset into manageable segments and processing only critical comparisons.
Solution Approach 2:
The patent applies partial action by computing edit distance not for all possible read pairs, but only for a subset of pairs involving representative reads. By selecting a small number of representative reads from each hash group and computing edit distances only for these representatives, the system achieves sufficient clustering completeness without the prohibitive cost of exhaustive pairwise comparisons.
3Productivity
If representative reads are designated to reduce computational expense, then processing efficiency is improved, but clustering accuracy may deteriorate
Solution Approach 1:
The patent incorporates feedback mechanisms where clustering results are iteratively refined. Representative reads are selected from hash groups, clustered using edit distance, and the results feed back into the process. The system can identify when representative reads may have been poorly selected and adjust subsequent selections accordingly, ensuring that efficiency gains from using representatives do not compromise overall clustering accuracy.
Data Source
Figure 1
Figure 2
Figure 3
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.