LCPAN: efficient variation graph construction using locally consistent parsing

Area of Science:

• Bioinformatics
• Computational Biology
• Genomic Data Analysis

Background:

• Genomic data is growing exponentially, demanding efficient string processing methods.
• Locally Consistent Parsing (LCP) partitions genome strings into consistent substrings ('cores') for compact representation.
• Existing sketching techniques can be less efficient in terms of representation size and analysis speed.

Purpose of the Study:

• To present the first iterative implementation of Locally Consistent Parsing (LCP) using LCPTOOLS.
• To introduce LCPAN, an efficient variation graph constructor.
• To demonstrate the performance improvements of LCPAN over existing tools like vg.

Main Methods:

• Developed an iterative implementation of Locally Consistent Parsing (LCP).
• Introduced LCPAN, a novel variation graph constructor.
• Benchmarked LCPAN against vg for speed and memory usage in variation graph construction.

Main Results:

• LCPAN generates variation graphs over 12 times faster than vg.
• LCPAN utilizes over 13 times less memory compared to vg.
• The iterative LCP implementation ensures consistent partitioning of genomic strings.

Conclusions:

• LCPTOOLS and LCPAN offer significant performance advantages for genomic data processing.
• These tools enable more efficient and scalable construction of variation graphs.
• The LCP approach provides a robust method for handling large-scale genomic datasets.