🙏 We thank all collaborating groups and contributing cohorts, for making this work possible! I would like to thank all co-authors, in particular @ludefranke.bsky.social and @urmovosa.bsky.social! (8/8)

🧩 These results provide an updated resource for studying human gene regulatory networks and for linking genetic variants to downstream molecular effects. Lead fine-mapped eQTLs and gene–gene links are available now! Genome-wide summary statistics will be made available upon publication. (7/8)

Colocalization with 87 GWAS traits suggests that trans-eQTLs can link additional disease-associated variants to genes. By integrating GWAS colocalization with gene–gene links, we can prioritize genes showing more-than-expected, consistent up- or down-regulation across traits. (6/8)

Cis-eQTLs for which we do observe downstream (trans) effects differ from those for which we don’t: they tend to map further from the TSS, are enriched for transcription factors, and are more likely to colocalize with complex‑trait loci. Understanding the mechanisms of this needs more work! (5/8)

The directed network is supported by external data, including Perturb‑seq, ChIP‑seq, and protein interaction datasets. (4/8)

We find 47,554 directed gene–gene regulatory links, inferred by integrating fine‑mapping with systematic cis–trans colocalization. Many upstream regulators, often transcription factors, give directional and dose‑dependent downstream effects. (3/8)

What’s new in eQTLGen Phase 2? Identification of 61,470 cis‑eQTLs and 48,953 trans‑eQTLs, with trans‑effects mapped at genome‑wide scale! For highly expressed genes we identify a strong negative relationship between genic evolutionary constraint and the number of cis-eQTLs. (2/8)