iMapSplice: Alleviating reference bias through personalized RNA-seq alignment.

Area of Science:

• Genomics
• Transcriptomics
• Bioinformatics

Background:

• Genomic variants impact gene splicing, affecting transcript function.
• Accurate RNA-sequencing (RNA-seq) read alignment across exon junctions is crucial for transcriptome profiling.
• Current alignment algorithms struggle with reads containing genomic variants, leading to biases and missed splice variants.

Purpose of the Study:

• To develop a novel algorithm, iMapSplice, for personalized mRNA transcriptome profiling.
• To improve RNA-seq read alignment accuracy in the presence of genomic variants.
• To enable the discovery of splice variants arising from splice site polymorphisms.

Main Methods:

• Developed iMapSplice, an algorithm utilizing personal genomic information for alignment.
• Performed unbiased alignment against genome indices containing both reference and alternative bases.
• Conducted comparative analyses using simulated and real RNA-seq datasets.

Main Results:

• iMapSplice demonstrated general improvements in read alignment and splice junction discovery.
• The algorithm significantly reduced allelic ratio biases in RNA-seq data.
• iMapSplice identified numerous previously uncharacterized splice junctions resulting from splice site polymorphisms.
• The method achieved these improvements with minimal computational overhead.

Conclusions:

• iMapSplice offers a robust solution for personalized transcriptome profiling by overcoming limitations of reference-based alignment.
• The algorithm effectively detects splice variants caused by splice site polymorphisms, enhancing the discovery of transcript diversity.
• iMapSplice provides a valuable tool for accurate RNA-seq analysis, particularly in personalized medicine and variant-rich samples.