Table 1 Main association results for variant-trait significant findings.

From: Association of protein function-altering variants with cardiometabolic traits: the strong heart study

Trait

Chr:position (hg19)

Gene

Marker exonic function

Amino acid change

Coded/Other allele

Minor Allele Count

N

Effect

Variance explained (%)

p-value

Functional prediction of SNV*

Serum creatinine

1:101,342,412

EXTL2

Missense

p.M148V

G/A

33

1125

 − 0.527

2.42

8.7 × 10–9

Deleterious

Serum creatinine

17:35,518,712

ACACA

Missense

p.P1683S

A/G

26

1125

 − 0.506

1.87

2.9 × 10–7

Deleterious

Fasting triglycerides

17:67,149,477

ABCA10

rs779392624 missense

p.G1369W

A/C

17

1124

 − 0.696

2.15

7.5 × 10–9

Deleterious

Fasting insulin

6:79,650,711

PHIP

Missense

p.T1722I

A/G

28

790

 − 0.369

1.30

2.1 × 10–6

Deleterious

Fasting insulin

9:73,152,248

TRPM3

rs760461668 missense

p.V1086M

A/G

185

792

0.166

1.70

4.8 × 10–8

Deleterious

Fasting insulin

11:18,637,366

SPTY2D1

rs756851199 missense

p.V152A

G/A

109

793

0.220

1.75

1.6 × 10–8

Neutral/tolerant

Fasting insulin

22:43,557,062

TSPO

rs566547284 missense

p.G63S

A/G

26

793

0.396

1.45

2.4 × 10–6

Deleterious

  1. For nonsynonymous rare variants and LOF variants, functional prediction algorithms were used to classify a SNV as deleterious based on agreement for at least three algorithms of prediction methods (see methods and Table S3). All SNVs listed in Table 1 had a CADD Phred score > 10–20, which is considered deleterious, except for rs756851199. Models adjusted for age, sex, center, and the first 10 principal components of ancestry. Fasting insulin was tested among non-diabetic individuals in models additionally adjusted for BMI. Amino acid change provided by the Variant Effect Predictor tool. N total number of participants. N/A, not available. Note three SNVs are not present in a publicly available database and lack rs#. Significance threshold p = 4.9 × 10–6 is based on number of SNVs and phenotypes tested.