AI Guided Screening Identifies Lily Moth Pheromone

Researchers from INRAE, Université Côte d’Azur and Nanjing Agricultural University have identified the major female sex pheromone component of the lily moth, Spodoptera picta, through an odorant receptor guided workflow. The peer reviewed study was published in BMC Biology on September 21, 2026, and combines computational prediction with chemical, physiological and behavioural validation rather than treating an artificial intelligence model as proof on its own.

Conventional pheromone discovery usually begins with compounds collected from live female insects. Researchers must know when pheromone production occurs, separate a complex chemical mixture and then determine which molecules actually affect male behaviour. That process can be difficult for species that are rare, hard to rear or subject to quarantine controls.

The team reversed the usual sequence. Genome analysis identified the lily moth versions of two candidate pheromone receptors, OR5 and OR75, known from related Spodoptera species. Artificial intelligence assisted structural modelling and molecular docking were then used to predict which volatile compounds could fit the receptors’ binding regions.

Laboratory testing supported the computational prediction. Functional receptor assays, gas chromatography and mass spectrometry of female gland extracts, antennal recordings and behavioural trials all pointed to (Z,E) 9,11 tetradecadienyl acetate as the major component. The compound produced the strongest antennal response and was sufficient to attract males under the study conditions.

The behavioural results were promising but came from small controlled assays. At a dose of 10 micrograms, 56.3 percent of 16 males showed wing fanning and 31.3 percent displayed hair pencils and contacted the source. A four compound synthetic blend produced those responses in 64.7 percent and 58.8 percent of 17 males respectively, while still performing less completely than natural gland extract.

Spodoptera picta is found in Asia and Oceania and its caterpillars damage lilies and related ornamental plants. It is not considered a pest of concern in mainland France. That distinction matters because the study establishes a discovery method and a candidate attractant for this particular moth, not a ready made control product for every agricultural pest.

The broader crop protection opportunity is a faster way to identify pheromones for current, emerging and invasive insects. Confirmed pheromones can support monitoring traps, mass trapping or mating disruption, potentially giving growers and public garden managers more targeted options where broad pesticide use is restricted.

Commercial use will require more work. Field trials must determine lure range, release rate, longevity, species specificity and performance across environments, and the weaker response to the synthetic blend suggests that trace components may still be missing. The advance is therefore a validated discovery workflow and a biologically active lead, not proof of field scale pest control.