TransMCL enables the assembly of full‐length coding genes for collecting complete hierarchical orthogroups

Abstract

Abstract Transcriptome sequencing technologies have revolutionized the field of phylogenomics by facilitating the identification of homologous genes for species without whole genome sequences. To infer complex evolution relationship among eukaryotes, it is essential to obtain complete sequences of protein‐coding genes to provide informative mutations. However, transcriptomes in eukaryotes often consist of a large number of duplicated genes and alternative isoforms, bringing great challenges for developing effective tools to obtain complete coding gene sequences. Here, we present a net‐flow based assembler TransMCL, which aims to assemble fragmented transcripts into complete sequences while eliminating redundant isoforms during homologue clustering. By employing Markov clustering strategies and homologous gene guidance, TransMCL can accurately assemble genes for multiple organisms in an affordable time frame, making it well‐suited for phylogenomic studies based on transcriptomic data. Our results demonstrate that TransMCL can assemble 89.95%–92.95% of the total expressed genes into near‐complete transcripts on benchmark plant/animal datasets. Furthermore, applying TransMCL to multiple transcriptomes in a single run enhances the completeness of genes, even in the absence of guidance homologues from closely related species. These findings highlight the potential of TransMCL in phylogenomic studies, enabling the comprehensive characterization of gene families at the whole genome scale. By overcoming the challenges of complex gene contents in eukaryotes, TransMCL can significantly enhance our understanding of evolution of gene families across species.