Systematic interrogation of the Conus marmoreus venom duct transcriptome with ConoSorter reveals 158 novel conotoxins and 13 new gene superfamilies
1 Division of Chemistry and Structural Biology, Institute for Molecular Bioscience, The University of Queensland, Brisbane, Qld 4072, Australia
2 Division of Genomics and Computational Biology, Institute for Molecular Bioscience, The University of Queensland, Brisbane, Qld 4072, Australia
BMC Genomics 2013, 14:708 doi:10.1186/1471-2164-14-708Published: 16 October 2013
Conopeptides, often generically referred to as conotoxins, are small neurotoxins found in the venom of predatory marine cone snails. These molecules are highly stable and are able to efficiently and selectively interact with a wide variety of heterologous receptors and channels, making them valuable pharmacological probes and potential drug leads. Recent advances in next-generation RNA sequencing and high-throughput proteomics have led to the generation of large data sets that require purpose-built and dedicated bioinformatics tools for efficient data mining.
Here we describe ConoSorter, an algorithm that categorizes cDNA or protein sequences into conopeptide superfamilies and classes based on their signal, pro- and mature region sequence composition. ConoSorter also catalogues key sequence characteristics (including relative sequence frequency, length, number of cysteines, N-terminal hydrophobicity, sequence similarity score) and automatically searches the ConoServer database for known precursor sequences, facilitating identification of known and novel conopeptides. When applied to ConoServer and UniProtKB/Swiss-Prot databases, ConoSorter is able to recognize 100% of known conotoxin superfamilies and classes with a minimum species specificity of 99%. As a proof of concept, we performed a reanalysis of Conus marmoreus venom duct transcriptome and (i) correctly classified all sequences previously annotated, (ii) identified 158 novel precursor conopeptide transcripts, 106 of which were confirmed by protein mass spectrometry, and (iii) identified another 13 novel conotoxin gene superfamilies.
Taken together, these findings indicate that ConoSorter is not only capable of robust classification of known conopeptides from large RNA data sets, but can also facilitate de novo identification of conopeptides which may have pharmaceutical importance.