The Bellerophon pipeline, improving de novo transcriptomes and removing chimeras

Area of Science:

• Bioinformatics
• Computational Biology
• Genomics

Background:

• Transcriptome quality control is crucial for RNA-Seq experiments.
• Assessing de novo assembled transcriptomes is difficult due to the absence of a reference genome.
• Chimeric sequences, formed by merging transcripts, can compromise transcriptome assembly quality.

Purpose of the Study:

• To develop and validate a method for assessing and improving the quality of de novo assembled transcriptomes.
• To focus on the identification and removal of chimeric sequences within transcriptome assemblies.
• To create a broadly applicable and user-friendly pipeline for transcriptome quality enhancement.

Main Methods:

• The Bellerophon pipeline integrates TransRate for quality assessment.
• It employs a transcripts per million (TPM) filter to remove lowly expressed contigs.
• CD-HIT-EST is utilized to eliminate highly identical contigs, addressing potential redundancy and chimeric artifacts.

Main Results:

• Bellerophon successfully removed 40% to 91.9% of chimeric sequences in benchmark transcriptome assemblies.
• The pipeline demonstrated a higher removal rate for chimeric contigs compared to non-chimeric ones.
• Validation involved computational chimera creation, identification in existing assemblies, and simulated RNA-Seq experiments.

Conclusions:

• The Bellerophon pipeline offers a robust solution for improving de novo transcriptome assemblies.
• Its sequence of filtration steps effectively addresses chimeric sequences, enhancing overall assembly quality.
• Bellerophon is broadly applicable and user-friendly, aiding researchers in obtaining more reliable transcriptome data.