FusionMap: detecting fusion genes from next-generation sequencing data at base-pair resolution

Area of Science:

• Genomics
• Bioinformatics
• Computational Biology

Background:

• Next-generation sequencing (NGS) enables fusion gene detection at transcriptomic and genomic levels.
• Current bioinformatics tools often overlook junction-spanning reads, relying heavily on paired-end approaches.
• Advancements in sequencing technology necessitate efficient aligners for longer reads to detect fusion events.

Purpose of the Study:

• To develop a novel bioinformatics tool, FusionMap, for accurate detection of fusion events.
• To enable direct alignment of junction-spanning single reads without prior knowledge of fusion regions.
• To improve the sensitivity and specificity of fusion gene detection using NGS data.

Main Methods:

• FusionMap aligns fusion reads directly to the genome.
• The method processes both single- and paired-end datasets from RNA-Seq and gDNA-Seq.
• FusionMap characterizes fusion junctions at base-pair resolution.

Main Results:

• FusionMap demonstrated high sensitivity and specificity in simulated RNA-Seq datasets.
• The tool outperformed paired-end approaches, especially with small inner distances between read pairs.
• Accurate fusion gene detection was validated in K562 chronic myeloid leukemia cell line using both RNA-Seq and gDNA-Seq data.

Conclusions:

• FusionMap offers an accurate and systematic solution for detecting fusion events using junction-spanning reads.
• The software integrates reference indexing, read filtering, alignment, and reporting.
• FusionMap is freely available for non-commercial use.