Evaluating the Performance of De Novo Assembly Methods for Venom-Gland Transcriptomics

Area of Science:

• Bioinformatics
• Genomics
• Molecular Biology

Background:

• Venom-gland transcriptomics is vital for studying venom evolution, ecology, function, and pharmacology.
• Accurate transcriptome assembly is essential for reliable inferences on venom variation.
• Current studies often rely on single assembly methods, potentially limiting recovery of complex gene families.

Purpose of the Study:

• To evaluate the performance of various de novo assembly methods for venom-gland transcriptomes.
• To assess the impact of k-mer size on transcript recovery.
• To identify optimal assembly strategies for recovering both toxin and nontoxin transcripts.

Main Methods:

• Evaluated multiple de novo transcriptome assembly software (e.g., Trinity, SDT, SeqMan NGen, Extender).
• Assembled transcriptomes from eleven snake and four scorpion venom glands.
• Varied k-mer sizes for specific assemblers to assess their impact on transcript recovery.

Main Results:

• Different assemblers perform optimally for toxin and nontoxin transcript recovery.
• SDT (small k-mers) and Trinity are best for nontoxin transcripts; SeqMan NGen + Extender excel for toxin transcripts.
• Extender uniquely assembled multiple isoforms of snake venom metalloproteinases, outperforming other methods.

Conclusions:

• Traditional assembly performance metrics do not predict toxin gene recovery.
• Combining results from multiple assemblers with diverse strategies is recommended for high-quality venom-gland transcriptomes.
• This approach enhances the recovery of complete and accurate toxin gene sets, crucial for functional studies.