Taxonomic identification of Morocco scorpions using MALDI-MS fingerprints of venom proteomes and computational modeling.

The venom of scorpions has been the subject of numerous studies. However, their taxonomic identification is not a simple task, leading to misidentifications. This study aims to provide a practical approach for identifying scorpions based on the venom molecular mass fingerprint (MFP). Specimens (251) belonging to fifteen species were collected from different regions in Morocco. Their MFPs were acquired using MALDI-MS. These were used as a training dataset to generate predictive models and a library of mean spectral profiles using software programs based on machine learning. The computational model achieved an overall recognition capability of 99 % comprising 32 molecular signatures. The models and the library were tested using a new dataset for external validation and to evaluate their capability of identification. We recorded an accuracy classification with an average of 97 % and 98 % for the computational models and the library, respectively. To our knowledge, this is the first attempt to demonstrate the potential of MALDI-MS and MFPs to generate predictive models capable of discriminating scorpions from family to species levels, and to build a library of species-specific spectra. These promising results may represent a proof of concept towards developing a reliable approach for rapid molecular identification of scorpions in Morocco. SIGNIFICANCE OF THE STUDY: With their clinical importance, scorpions may constitute a desirable study model for many researchers. The first step in studying scorpion is systematically identifying the species of interest. However, it can be a difficult task, especially for the non-experts. The taxonomy of scorpions is primarily based on morphometric characters. In Morocco, the high number of species and subspecies mainly endemic, and the morphological similarities between different species may result in false identifications. This was observed in many reports according to the scorpion experts. In this study, we describe a reliable practical approach for identifying scorpions based on the venom molecular mass fingerprints (MFPs). By using two software programs based on machine learning, we have demonstrated that these MFPs contains sufficient inter-specific variation to differentiate between the scorpion species mentioned in this study with a good accuracy. Using a drop of venom, this new approach could be a rapid, accurate and cost saving method for taxonomic identification of scorpions in Morocco.