Hybrid assembly of the African savanna elephant whole blood transcriptome using short- and long-read data

Background Understanding of immune responses and biomarker discovery in African savanna elephants ( Loxodonta africana ) has been limited due to the paucity of transcriptomic resources. Therefore, this study aimed to generate a high-quality whole blood transcriptome for this species using both short- and long-read sequencing data and develop a workflow for transcriptome assembly in non-model species. To determine the highest quality assembly, multiple transcriptome assemblies were generated and evaluated using combinations of short- and long-read sequencing data, together with reference-guided ( StringTie, StringTie2 and StringTie2 annotation-guided) and de novo ( rnaSPAdes , Trinity and RNA-Bloom2 ) assembly approaches. Results Paired-end short-read next-generation sequencing data were generated using the MGI platform, while long-read sequencing data were produced using the third-generation Oxford Nanopore Technologies platform. From the assembly tools compared in this study, StringTie2 with annotation-guided assembly produced the highest-quality transcriptome. This assembly exhibited high transcript continuity, low levels of unannotated transcripts, extensive functional annotation, and a BUSCO completeness score of 98.4%. Functional characterisation represented diverse biological processes, molecular functions, cellular components, and pathways relevant to immune and cellular regulation. Conclusion These findings support the use of this whole blood transcriptome as a reference resource for future studies related to disease in African savanna elephants. The transcriptome and associated annotations provide a foundation for future gene expression, functional genomics, and genome annotation research in this endangered species.