Genome-wide association studies (GWAS) have identified thousands of genetic loci associated with blood cell traits. However, the identification and interpretation of candidate causal variants and genes remain challenging. With increasing power to detect genetic associations, many loci reveal multiple statistically independent signals, which - while further adding complexity - also include rare variants with large effect sizes that may guide functional validation. This thesis aims to develop a framework for systematically prioritising candidate functional variants and genes from GWAS and to experimentally elucidate their mechanistic consequences using a tractable model, as provided by the haematopoietic system.

A recent GWAS of 36 blood cell traits identified 1399 genetic variants associated with 12 red blood cell traits. I implemented a systematic bioinformatics strategy to prioritise candidate functional variants and genes for experimental follow-up by leveraging variant allele frequency and consequence prediction, as well as cell type-specific gene expression and epigenomic information. This approach identified 17 loci with independent signals in protein coding and putative regulatory regions, with candidate genes implicated in erythroid differentiation (TRIM10), cytoskeletal structure (ANK1, SPTA1, TLN2), mechanosensitive ion channel activity (PIEZO1, TMC8 ), sphingolipid signalling (PLD1, SPHK1, S1PR2), iron homeostasis (TF, TFR2) and nucleotide metabolism (GMPR, PKLR).

To uncover the function of candidate genes in relation to red blood cell development, I established protocols to generate CRISPR/Cas9-based model systems in human erythroid cell lines (K562 and HUDEP-2) and induced pluripotent stem cells (iBOB). As a proof-of-principle, I assessed the consequences of mono- and biallelic GMPR disrupting mutations on cellular proliferation, erythroid differentiation and gene expression in iPS-derived erythroblasts. Complementing the experiments in cellular models, I designed and conducted a recall-by- genotype study in healthy volunteers carrying putative functional GMPR variants. Applying mass spectrometry, isolated erythrocytes from recalled participants with rare coding variants in GMPR had lowered GMPR protein levels and showed increased expression of Ras-related GTPases. These findings shed light on the important role of GMPR in purine metabolism and erythropoiesis.

My thesis demonstrates how combining a systematic annotation strategy of GWAS variants with targeted experimental investigation can identify candidate variants and genes likely to be causally implicated in complex traits and elucidate the underlying molecular mechanism.