Premium
Genotype‐based hierarchical clustering reveals a panel of polymorphisms in one carbon metabolism that are associated with obesity
Author(s) -
Corbin Karen D,
Spencer Melanie D,
Costa Kerry-Ann,
Sha Wei,
Abdelmalek Manal F,
Pan Yiping,
Suzuki Ayako,
Guy Cynthia D,
Cardona Diana M,
Torquati Alfonso,
Diehl Anna Mae,
Zeisel Steven H
Publication year - 2012
Publication title -
the faseb journal
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.709
H-Index - 277
eISSN - 1530-6860
pISSN - 0892-6638
DOI - 10.1096/fasebj.26.1_supplement.819.18
Subject(s) - single nucleotide polymorphism , methylenetetrahydrofolate reductase , genetics , body mass index , biology , genome wide association study , genotype , obesity , bioinformatics , medicine , endocrinology , gene
We previously demonstrated that single nucleotide polymorphisms (SNPs) in the phosphatidylcholine (PtdCho) flippase ABCB4 were associated with obesity and proposed that PtdCho metabolism is a key player in energy metabolism. We sought to identify patterns of SNPs in one carbon metabolism, and related pathways, that are associated with obesity by utilizing hierarchical clustering to group people by genotype rather than phenotype [n = 720; body mass index (BMI) range 17–76]. Clusters were generated by selecting the top SNPs associated with BMI based on a linear regression model (n = 99 out of 260; cutoff raw p‐value < 0.1) and removing highly correlated SNPs (R 2 ≥ 0.95). The final list included 67 SNPs. The 9 clusters generated were characterized as related to BMI, gender, ethnicity, diabetes diagnosis, hepatic steatosis, and age. We found 3 high and 1 low BMI clusters. We identified SNPs in ABCB4, BHMT, PEMT, FADS2, MTHFD1, MTHFR, SCD, and SLC44A1 that were significantly different in the heavy versus lean clusters. Additional phenotypes were different in the heavy versus lean clusters, particularly gender and ethnicity. This methodology may expand our understanding of the concurrent genetic patterns associated with the diverse obese phenotype and allow us to find relevant genetic factors in specific groups of people that may be missed with standard statistical approaches. This work is supported by NIH grant DK055865.