Younis, InasYKhattab, Amira RSelim, Nabil MSobeh, MansourElhawary, Seham SEl Bishbishy, Mahitab H2022-03-282022-03-2823/03/2022https://doi.org/10.1038/s41598-022-08479-4http://repository.msa.edu.eg/xmlui/handle/123456789/4898Seven avocado “Persea americana” seeds belonging to 4 varieties, collected from diferent localities across the world, were profled using HPLC–MS/MS and GC/MS to explore the metabolic makeup variabilities and antidiabetic potential. For the frst time, 51 metabolites were tentatively- identifed via HPLC–MS/MS, belonging to diferent classes including favonoids, bifavonoids, naphthodianthrones, dihydrochalcones, phloroglucinols and phenolic acids while 68 un-saponifed and 26 saponifed compounds were identifed by GC/MS analysis. The primary metabolic variabilities existing among the diferent varieties were revealed via GC/MS-based metabolomics assisted by unsupervised pattern recognition methods. Fatty acid accumulations were proved as competent, and varietal-discriminatory metabolites. The antidiabetic potential of the diferent samples was explored using in-vitro amylase and glucosidase inhibition assays, which pointed out toGwen (KG) as the most potent antidiabetic sample. This could be attributed to its enriched content of poly-unsaturated fatty acids and polyphenolics. Molecular docking was then performed to predict the most promising phytoligands in KG variety to be posed as antidiabetic drug leads. The highest in-silico α-amylase inhibition was observed with chrysoeriol-4′-O-pentoside-7-O-rutinoside, apigenin-7-glucuronide and neoeriocitrin which might serve as potential drug leads for the discovery of new antidiabetic remedies.en-USDiseasesDrug screeningHigh-throughput screeningTarget identificationArticlehttps://doi.org/10.1038/s41598-022-08479-4