A newly published paper shows that a local photoexcitation of FAD as the initial step in cryptochrome-based magnetoreception is insufficient to explain green-light-activated magnetoreception. Even when embedded inside a cryptochrome protein matrix, FAD has no absorbance in the green-light range of the visual part of the electromagnetic spectrum.
This result leads to the conclusion that there must be "something else" involved in generation of the well-established radical pair in cryptochrome proteins, if they should be generated under green light, and this "something else" could very well be one of the secondary photoreceptors that have been already been discovered in some cryptochrome species (e.g. MTHF).
The study is based on advanced computational methods and studies the absorption properties of six different cryptochrome structures, i.e. from Arabidopsis thaliana (plant), Drosophila melanogaster (fruitfly), Mus Musculus (mouse), Xenopus laevis (frog) and both cryptochrome 1 and 4 from Erithacus rubecula (european robin).