ChiTaH: a fast and accurate tool for identifying known human chimeric sequences from high-throughput sequencing data

Area of Science:

• Genomics
• Bioinformatics
• Cancer Biology

Background:

• Fusion genes, or chimeras, arise from joined sequences of different genes and can drive cancer development.
• Accurate identification of these driver fusions is crucial for cancer diagnosis and targeted therapies.
• Existing computational tools for chimera detection have limitations in sensitivity, specificity, and quantification.

Purpose of the Study:

• To develop a novel, reference-based computational approach for accurate identification of chimeric transcripts.
• To evaluate the performance of the new method against existing chimera detection tools.
• To demonstrate the utility of the new method in uncovering chimera heterogeneity.

Main Methods:

• Development of ChiTaH (Chimeric Transcripts from High-throughput sequencing data), a reference-based method utilizing a database of 43,466 known human chimeras.
• Benchmarking ChiTaH against four other chimera identification methods using simulated and real sequencing datasets (DNA-Seq, RNA-Seq).
• Experimental validation of findings, including uncovering chimera heterogeneity in the K-562 cell line.

Main Results:

• ChiTaH demonstrated superior accuracy and speed in identifying known human chimeras compared to other tested methods.
• The method effectively processed both simulated and real-world high-throughput sequencing data.
• ChiTaH revealed heterogeneity of the BCR-ABL1 chimera in bulk and single-cell K-562 samples, confirmed experimentally.

Conclusions:

• ChiTaH represents a significant advancement in the accurate and efficient detection of chimeric transcripts from sequencing data.
• The reference-based approach overcomes limitations of existing methods, improving chimera identification for cancer research.
• ChiTaH's ability to detect chimera heterogeneity offers new insights into cancer biology and potential therapeutic strategies.