Magic-BLAST, an accurate RNA-seq aligner for long and short reads

Area of Science:

• Genomics
• Bioinformatics

Background:

• Next-generation sequencing (NGS) generates millions of reads, necessitating efficient alignment tools.
• Accurate intron discovery and long-read alignment from RNA data remain challenging for existing software.

Purpose of the Study:

• Introduce Magic-BLAST, a novel aligner designed for accurate RNA-genome alignment and intron discovery.
• Evaluate Magic-BLAST's performance against established aligners using diverse datasets.

Main Methods:

• Developed Magic-BLAST with optimized spliced alignment scoring and selective masking during seed selection.
• Tested on real-world RNA-seq data from PacBio, Roche, and Illumina platforms.
• Utilized six benchmark datasets and idealized human mRNA sequences for evaluation.

Main Results:

• Magic-BLAST demonstrates superior intron discovery across various conditions.
• Achieved best performance in mapping reads longer than 250 bases from all platforms.
• Showed robustness to high mismatch rates and extreme base compositions.

Conclusions:

• Magic-BLAST is the leading tool for intron discovery and long-read mapping in RNA sequencing.
• Its versatility, speed, and robustness make it suitable for diverse genomic applications.
• Supports alignment against BLAST databases or FASTA files, and accepts FASTQ input or SRA accessions.