For the genome-wide suggestive and significant loci, we explored potential causative genes at each locus using a strategy that combined publicly available datasets and prioritization tools and eQTLs from peripheral blood– and immune cell type–specific datasets.

Coding variation. To obtain the putative functional effect of associated variants, significantly or suggestively associated SNPs were annotated with VEP (11) in May 2016. Because of alternative splicing, each gene can have more than one transcript, and, consequently, depending on the transcript, one SNP can have a different effect on protein function.

SMR/HEIDI analysis for pleiotropy with gene expression. To test for potential pleiotropy between gene expression and IgG glycosylation, we performed SMR with HEIDI analysis, developed by Zhu et al. (12). SMR analysis provides evidence for pleiotropy but cannot distinguish if the associations are driven by the same or highly correlated but distinct causal variants. The subsequent HEIDI test allowed us to distinguish pleiotropy from LD. The test was performed on a publicly available peripheral blood dataset from Westra et al (14) and immune cell–specific gene expression from the CEDAR dataset (13). Among others, this dataset contains expression in B lymphocytes (CD19), helper T lymphocytes (CD4), cytotoxic T cells (CD8), macrophages (CD14), neutrophils (CD15), and platelets (PLA).

We set a threshold for the SMR test at PSMR ≤ 1.9 × 10−5, corresponding to a Bonferroni correction for 2622 tests, number of regions where genome-wide significant top regional IgG glycan SNP (or its proxy) was also available in any of the gene expression associations. All regions with significant PSMR and PHEIDI ≥ 0.05 were considered to exhibit concordance in regional association patterns and therefore showed evidence of sharing the underlying unobserved causal variant. For these regions, we can suggest that they are pleiotropic. Details of the algorithm can be seen in the Supplementary Note.

Prioritization using DEPICT. DEPICT (10) was run on the merged list of independent SNPs obtained from the GCTA-COJO analysis to identify gene sets enriched for genes near associated variants. Suggestive independent SNPs (P ≤ 5 × 10−8) were submitted to DEPICT, release 194 (10). The list of independent SNPs was created by merging glycan-wise GCTA-COJO results, resulting in 113 SNPs. Given that different glycans can have a different lead SNP (but in high LD), this list was additionally pruned by applying PLINK clumping (43), where all SNPs within 500 kb and LD R2 > 0.1 of the SNP with the strongest association were assigned to the same clump. LD was estimated using 1000 Genomes Project Phase 1 CEU [Utah Residents (CEPH) with Northern and Western European Ancestry], GBR (British in England and Scotland), and TSI (Toscani in Italy) data.

Gene set enrichment analysis. GO enrichment analysis was performed using FUMA GENE2FUNC (15) analysis based on MSigDB c5 with default parameters and All genes as background genes. Glycosylation-related gene sets were defined as any GO gene set whose description contained words “glyc,” “sacch. fucose,” “carbo,” or “hexose.” Immune system–related gene sets were defined in the same manner, but searching for words “immune,” “B_CELL,” “lymphocyte,” “leukocyte,” “T_CELL,” “hemopoi,” and “myeloid,” while transcription-related gene-sets were defined using the keywords “transcription” or “expression.” DEPICT pathways and tissue enrichments analyses were performed as described above.

Note: The content above has been extracted from a research article, so it may not display correctly.

Please log in to submit your questions online.
Your question will be posted on the Bio-101 website. We will send your questions to the authors of this protocol and Bio-protocol community members who are experienced with this method. you will be informed using the email address associated with your Bio-protocol account.

We use cookies on this site to enhance your user experience. By using our website, you are agreeing to allow the storage of cookies on your computer.