This article explores the development of a linguistic database for the translation of Uzbek grammatical forms into English. The study investigates translation equivalence and divergence at the grammatical level through the lens of contrastive linguistics and computational linguistics. Affixal forms, auxiliary verbs, modal expressions, and syntactic structures in Uzbek are systematically matched with their grammatical and semantic equivalents in English. Based on this mapping, a modern linguistic platform named UzEngGram.uz has been developed. The paper outlines stages of grammatical identification, classification, translation modeling using a parallel corpus, automation through NLP tools, and the creation of a lexical-grammatical database. It is further argued that the resulting resource serves as a critical tool for machine translation systems, linguistic analysis, and educational applications.