Sim4cc: a cross-species spliced alignment program.

Area of Science:

• Genomics
• Bioinformatics
• Computational Biology

Background:

• Next-generation sequencing rapidly generates new genomes, outpacing gene annotation resources.
• Comparing existing cDNA sequences from related species is a common method for annotating novel genomes.
• Existing spliced alignment tools lack the sensitivity and accuracy needed for interspecies comparisons.

Purpose of the Study:

• To develop a novel spliced alignment tool, sim4cc, to overcome limitations of existing methods for interspecies genome annotation.
• To enhance the accuracy and sensitivity of gene model prediction in newly sequenced genomes.

Main Methods:

• Developed sim4cc, incorporating universal spaced seeds for increased sensitivity across evolutionary distances.
• Integrated powerful splice signal models and evolutionarily-aware alignment techniques to improve gene model accuracy.
• Tested sim4cc on vertebrate comparisons at diverse evolutionary distances.

Main Results:

• Sim4cc demonstrated significantly higher sensitivity than existing alignment programs in interspecies comparisons.
• Achieved over 10% higher sensitivity than the closest competitor in some vertebrate comparisons.
• Maintained comparable speed to its predecessor, sim4, while offering superior performance.
• Successfully mapped 64,000 Fagus grandifolia ESTs and unigenes to the poplar genome, demonstrating high-throughput capability.

Conclusions:

• Sim4cc is a highly sensitive and accurate tool for interspecies spliced alignment, crucial for annotating newly sequenced genomes.
• The tool's features enable effective gene model prediction and can be integrated into high-throughput annotation pipelines.
• Sim4cc advances comparative genomics by facilitating efficient gene discovery across diverse species.