A scalable and memory-efficient algorithm for de novo transcriptome assembly of non-model organisms.

Area of Science:

• Bioinformatics
• Genomics
• Computational Biology

Background:

• De novo assembly algorithms enable transcriptome studies in non-model organisms.
• Existing transcriptome assembly tools are memory-intensive, limiting their use to small datasets.

Purpose of the Study:

• To develop a memory-efficient transcriptome assembly algorithm for large RNA-Seq datasets.
• To enable the study of entire transcriptomes from non-model organisms.

Main Methods:

• Developed a novel transcriptome assembly algorithm utilizing computing clusters with moderate memory.
• Implemented techniques to minimize memory consumption during assembly.

Main Results:

• The algorithm recovers alternatively spliced isoforms and expression levels.
• It efficiently processes large RNA-Seq datasets (hundreds of gigabases).
• Achieved comparable or improved accuracy over existing algorithms.

Conclusions:

• The new strategy minimizes memory usage for transcriptome assembly.
• It supports incremental updates with new sequencing libraries.
• Facilitates comprehensive transcriptome analysis for non-model organisms.