I’ve been investigating the Variant-to-Gene (V2G) assignments in Open Targets using GraphQL API (using otargen package) . I’ve noticed an intriguing pattern: for certain gene targets associated with variants, the ‘typeId’ is reported as ‘NA’, and the Open Targets website omits the distance information for these associations. Upon further examination, I’ve observed that these particular gene targets are generally located more than 500 kb away from their associated variants.
Given that these associations lack supporting evidence from quantitative trait loci (QTLs) or chromatin interaction data, I’m curious about the rationale behind reporting them as targets without distance information. Shouldn’t the distance itself serve as a ‘typeId’ in cases where no other corroborating data exists?
Additionally, I’m interested in understanding the criteria Open Targets uses for including these long-range (>500 kb) associations in their database, especially when they diverge from the more common proximity-based assignments. Could you provide insights into the methodology and decision-making process behind these particular V2G assignments?