GRASShopPER-An algorithm for de novo assembly based on GPU alignments.

Area of Science:

• Genomics
• Bioinformatics
• Computational Biology

Background:

• Next-generation sequencing (NGS) generates vast amounts of short DNA reads, posing significant computational challenges for *de novo* genome assembly.
• Accurate genome reconstruction from short reads is crucial for various biological and medical applications.

Purpose of the Study:

• To develop and evaluate GRASShopPER, a novel genome assembler designed to overcome the computational hurdles of *de novo* assembly from short NGS reads.
• To improve the accuracy and contiguity of genome assemblies.

Main Methods:

• GRASShopPER employs an overlap-layout-consensus strategy.
• It utilizes efficient GPU implementation for sequence alignment in graph construction.
• A two-part fork detection method with a greedy hyper-heuristic algorithm is used for repeat identification and reconstruction.

Main Results:

• GRASShopPER was evaluated on bacterial, nematode, and human chromosome datasets using QUAST.
• The assembler produced contigs with the largest genome coverage compared to other assemblers.
• GRASShopPER maintained competitive NG50 values and low misassembly rates.

Conclusions:

• GRASShopPER demonstrates superior performance in *de novo* genome assembly from short NGS reads.
• Its efficient algorithms and GPU implementation enable accurate and contiguous genome reconstruction.
• The tool shows promise for advancing genomic research by providing high-quality assemblies.