Table 2. Proposed molecular consequences at multi-signal loci.
| Signal | Lead variant and adiponectin-decreasing allele | Number of candidate variants (r2≥0.80) | Candidate functional variant(s) based on allelic differences in experimental assays | Candidate transcript and direction based on colocalized eQTL | Putative molecular consequence |
|---|---|---|---|---|---|
| ADIPOQ | |||||
| A | rs199938283-C | 15 | rs76071583 | ↓ ADIPOQ, ↓ LINC02043 | Decreased ADIPOQ and LINC02043 enhancer activity; decreased transcriptional activity (Fig 2) |
| B | rs4632532-C | 14 | - | - | Assumed decreased ADIPOQ enhancer activity |
| C | rs16861209-C | 12 | - | - | Assumed decreased ADIPOQ enhancer activity |
| D | rs73187787-C | 7 | rs13322149, rs55958900 | - | Increased distal enhancer ADIPOQ-AS1 activity; increased transcriptional activity (Fig 2) |
| E | rs17366653-C | 1 | rs17366653 | - | ADIPOQ splice variant; reported to increase the proportion of ADIPOQ isoform lacking exon 2 and 3 [22] |
| F | rs17846866-G | 2 | - | ↑ ADIPOQ-AS1 | Assumed increased ADIPOQ-AS1 enhancer activity |
| G | rs62625753-A | 1 | rs62625753 (Gly90Ser) | - | ADIPOQ missense variant (G90S); reported to decrease adiponectin multimerization[21] |
| CDH13 | |||||
| A | rs12051272-T | 8 | rs12051272 | ↑ CDH13 | Decreased CDH13 proximal intron 1 promoter activity (Fig 4, S17 Fig) |
| B | rs4782722-G | 6 | rs4782722 | - | CDH13 distal intron 1 enhancer activity; increased transcriptional activity (Fig 4, S17 Fig) |
For candidate transcripts and putative molecular consequences, the directions of effect are based on the allele associated with decreased adiponectin levels in the METSIM GWAS. The putative molecular consequence column includes allele-specific functional evidence from this work and other cited publications. When no direct evidence exists for molecular consequences of a variant (CDH13 signal ‘A’ and ADIPOQ signals ‘B’, ‘C’, and ‘F’), we proposed molecular functions and candidate transcript based on candidate variant locations.