Artificial neural networks & cuticular hydrocarbons for the identification of the developmental stages of Sarcophaga peregrina (Diptera: Sarcophagidae)

Sarcophaga peregrina (Robineau-Desvoidy, 1830) (Diptera: Sarcophagidae), a synanthropic flesh fly species found in different parts of the world, is of medical and forensic importance. Morphological identification of the developmental stages of S. peregrina is currently restricted due to the low number of thoroughly trained professionals. Here, we combined artificial neural networks with cuticular hydrocarbons of S. peregrina to explore a new method for stage identification. Here, we studied the cuticular hydrocarbon profile of S. peregrina of the whole life cycle from larval stage to adult stage by GC-MS. Sixty-three compounds with carbon chain length ranging from 8 to 36 were detected totally, which could be classified into four classes: n-alkanes, branch alkanes, alkenes, and unknowns. Three-layers neural networks were used to cluster, with over 80% accuracy, as the validation set shown. Our researchers find that ANN is an important method for autonomous developmental stage identification. Expanding the data set of cuticular hydrocarbons makes it possible to predict developmental timing, which could be a potential method for postmortem interval estimation.