***************************************************
* Genome-wide association results                 *
* with targeted and non-targeted metabolic traits *
* in urine of SHIP-0 cohort participants.         *
***************************************************

All metabolic traits were determined in one-dimensional 1H NMR spectra.
The targeted traits were derived by manual spectral annotation using Chenomx Profiler 7.0. 
For the non-targeted traits (NMR chemical shifts), the spectra were automatically aligned and feature reduced using the FOCUS package.
All metabolic traits were normalized to the annotated creatinine concentrations do compensate for dilution effects. For both targeted and non-targeted traits, pairwise ratios were calculated. All metabolic traits were log10 transformed prior to analysis. Only traits with at least 300 valid data points were tested for genetic association using age- and gender-corrected linear models (under the assumption of an additive effect) in PLINK v. 1.07.

For details, see the Methods section of the paper.

The results file contain the association results for both autosomal genotyped variants (Affymetrix Human SNP Array 6.0) and, in selected regions, imputed variants (1000 genomes imputation).

For associations to single metabolic traits, we list all associations with P < 0.05.
In case of associations to ratios, we list all associations with P < 1e-4 and P-Gain >= 10.

Note that some associations have P values below 1.0e-308. Since these P values are below the double-precision floating-point limit, they are listed as "0" (additionally, in case of ratios the PGAIN is "Inf").


Columns description
===================

CHR	Chromosome the variant is located on (according to GRCh37).
POS	Chromosomal position of the variant in bp (according to GRCh37).
SNP	rs identifier of genetic variant.
EA	Effect allele (= minor allele of variant)
MET(1,2)	(targeted traits only) metabolite name according to the Chenomx Profiler. In case of ratios, MET1 is the metabolite that exhibits the stronger association signal to the genetic variant.
MET(1,2)ID	(targeted traits only) Chenomx ID for this metabolite. See mapping table below for the HMDB, KEGG Compound, and CAS IDs.
PPM(1,2)	(non-targeted traits only) chemical shift in p.p.m. for the tested targeted trait(s). In case of ratios, PPM1 is the trait that exhibits the stronger association signal to the genetic variant.
P	P-value of the association (corrected for age and gender)
PGAIN 	(ratios only) P(MET1)/P(MET1/MET2) (targeted traits); P(PPM1)/P(PPM1/PPM2) (non-targeted traits)
BETA	slope of the linear model
BETAPRIME	(10^BETA) – 1
SE	standard error (computed as BETA/STAT, where STAT is "t-statistics" in the PLINK output)
NMISS	number of complete genotype-metabolic trait pairs used for the association test



Mapping of Chenomx IDs to KEGG, HMDB, and CAS
=============================================

Metabolite	Chenomx	KEGG	HMDB	CAS
acetate	CNX0009	C00033	HMDB00042	64-19-7
acetone	CNX0011	C00207	HMDB01659	67-64-1
alanine	CNX0232	C00041	HMDB00161	56-41-7
asparagine	CNX0231	C00152	HMDB00168	70-47-3
betaine	CNX0015	C00719	HMDB00043	107-43-7
carnitine	CNX0021	C00318	HMDB00062	541-15-1
choline	CNX0302	C00114	HMDB00097	62-49-7
cis-aconitate	CNX0351	C00417	HMDB00072	585-84-2
citrate	CNX0022	C00158	HMDB00094	77-92-9
creatine	CNX0023	C00030	HMDB00064	57-00-1
dimethylamine	CNX0027	C00543	HMDB00087	124-40-3
ethanolamine	CNX0097	C00189	HMDB00149	141-43-5
ethanol	CNX0205	C00469	HMDB00108	64-17-5
formate	CNX0032	C00058	HMDB00142	64-18-6
glucose	CNX0183	C00031	HMDB00122	50-99-7
glycine	CNX0242	C00037	HMDB00123	56-40-6
glycolate	CNX0103	C00160	HMDB00115	79-14-1
hippurate	CNX0037	C01586	HMDB00714	495-69-2
histidine	CNX0217	C00135	HMDB00177	71-00-1
hypoxanthine	CNX0164	C00262	HMDB00157	68-94-0
isopropanol	CNX0072	C01845	HMDB00863	67-63-0
lactate	CNX0042	C00186	HMDB00190	79-33-4
lysine	CNX0291	C00047	HMDB00182	56-87-1
maleate	CNX0416	C01384	HMDB00176	110-16-7
mannitol	CNX0201	C00392	HMDB00765	69-65-8
methanol	CNX0178	C00132	HMDB01875	67-56-1
methylamine	CNX0044	C00218	HMDB00164	74-89-5
N,N-dimethylglycine	CNX0028	C01026	HMDB00092	1118-68-9
O-acetylcarnitine	CNX0195	C02571	HMDB00201	3040-38-8
O-phosphocholine	CNX0321	C00588	HMDB00284	107-73-3
propylene glycol	CNX0262	C00583	HMDB01881	57-55-6
pyroglutamate	CNX0119	C01879	HMDB00267	98-79-3
succinate	CNX0060	C00042	HMDB00254	110-15-6
sucrose	CNX0193	C00089	HMDB00258	57-50-1
tartrate	CNX0450	C00898	HMDB00956	87-69-4
taurine	CNX0061	C00245	HMDB00251	107-35-7
threonine	CNX0219	C00188	HMDB00167	72-19-5
trans-aconitate	CNX0275	C02341	HMDB00958	4023-65-8
trigonelline	CNX0334	C01004	HMDB00875	535-83-1
trimethylamine-N-oxide	CNX0063	C01104	HMDB00925	1184-78-7
tyrosine	CNX0243	C00082	HMDB00158	60-18-4
uracil	CNX0139	C00106	HMDB00300	66-22-8
urea	CNX0064	C00086	HMDB00294	57-13-6
valine	CNX0215	C00183	HMDB00883	72-18-4
1,6-anhydro-β-D-glucose	CNX0261	NA	HMDB00640	498-07-7
1-methylnicotinamide	CNX0338	C02918	HMDB00699	3106-60-3
2-hydroxyisobutyrate	CNX0069	NA	HMDB00729	594-61-6
3-aminoisobutyrate	CNX0091	C05145	HMDB03911	144-90-1
3-hydroxyisovalerate	CNX0284	NA	HMDB00754	625-08-1
3-hydroxymandelate	CNX0363	NA	HMDB00750	17119-15-2
3-indoxylsulfate	CNX0006	NA	HMDB00682	487-94-5
4-hydroxyphenylacetate	CNX0007	C00642	HMDB00020	156-38-7
xylose	CNX0089	C00181	HMDB00098	58-86-6
π-Methylhistidine	CNX0076	C01152	HMDB00479	368-16-1
τ-Methylhistidine	CNX0001	NA	HMDB00001	332-80-9

