Table 1.

Model parameters for phenotypes, case-control (upper section) and quantitative (lower section)

Phenotype	π1	${\mathit{\sigma }}_{\mathit{b}}^{\mathit{2}}$	${\mathit{\sigma }}_{\mathit{c}}^{\mathit{2}}$	S	${\mathit{p}}_{\mathit{c}\mathit{1}}$	mc	wc	${\mathit{\sigma }}_d^2$	${\mathit{p}}_{d1}$	md	wd	${\sigma }_0^2$
SCZ 2014	5.28e−2	1.4e−6	3.6e−5	−0.52	0.07	87	352	1.4e−4	5.2e−3	519	7	1.07
BIP	5.91e−2	1.2e−6	4.5e−5	−0.40	6.7e−2	102	414	—–	—–	—–	—–	1.01
CD	9.55e−4	5.0e−5	5.5e−4	−0.64	0.20	176	604	7.6e−2	1.0e−4	—–	—–	1.14
UC	1.16e−3	3.6e−5	4.0e−4	−0.67	0.16	173	627	8.0e−2	1.0e−4	—–	—–	1.12
CAD	1.88e−3	1.1e−5	9.2e−5	−0.51	6.9e−2	171	683	5.3e−3	3.6e−4	102	7	0.92
AD Chr19	4.34e−4	1.0e−4	6.1e−3	−0.57	0.73	35	89	—–	—–	—–	—–	1.09
AD NoC19	1.05e−3	1.8e−5	2.6e−4	−0.52	2.9e−2	264	6	—–	—–	—–	—–	1.04
ALS Chr9	1.12e−2	7.3e−6	3.9e−3	−0.01	1.8e−3	106	6	—–	—–	—–	—–	0.99
Edu	1.43e−2	1.7e−6	7.8e−6	−0.44	0.45	111	339	8.5e−5	6.3e−3	441	7	0.94
IQ 2018	1.27e−2	7.5e−7	6.2e−6	−0.51	0.38	122	309	3.6e−5	7.8e−2	561	6	1.17
Height 2010	1.02e−3	4.1e−5	2.0e−4	−0.44	0.20	322	1243	—–	—–	—–	—–	0.90
Height 2014	1.15e−3	3.7e−5	1.6e−4	−0.46	0.21	242	929	—–	—–	—–	—–	1.57
Height 2018	2.50e−3	8.7e−6	8.9e−5	−0.43	0.37	210	739	—–	—–	—–	—–	2.12
HDL	2.54e−3	1.1e−5	4.5e−4	−0.79	1.4e−2	143	599	2.2e−2	1.1e−3	66	7	0.91
LDL	5.84e−3	3.3e−6	2.4e−4	−0.52	8.8e−3	336	1417	7.3e−3	2.2e−4	346	6	0.92
BMI GIANT 2015	1.54e−3	2.2e−5	4.5e−4	0.00	4.4e−3	288	12	—–	—–	—–	—–	0.85
BMI 2018	2.34e−3	1.6e−5	3.0e−4	0.11	3.8e−3	268	8	—–	—–	—–	—–	1.72
TC	1.15e−3	1.7e−5	6.2e−4	−0.97	2.1e−2	140	583	2.9e−4	3.4e−2	—–	—–	0.92

π ₁ is the overall proportion of the 11 million SNPs from the reference panel that are estimated to be causal. $p_c(L⩾1)$ is the prior probability multiplying the “c” Gaussian, which has variance ${\sigma }_c^2{H}^S$, where H is the reference SNP heterozygosity. Note that $p_c(L)$ is just a sigmoidal curve, and can be characterized quite generally by three parameters: the value $p_{c1}\equiv p_c(1)$ at L = 1; the total LD value L = m_c at the mid-point of the transition, i.e. $p_c(m_c)=p_{c1}/2$ (see the middle gray dashed lines in Figure 1, which shows examples of the function $p_c(L)$); and the width w_c of the transition, defined as the distance (in L) between where the curve falls to 95% and 5% of $p_{c1}$ (distance between the flanking red dashed lines in Figure 1). Note that for AD Chr19, AD NoC19, and ALS Chr9, π₁ is the fraction of reference SNPs on chromosome 19, on the autosome excluding chromosome 19, and on chromosome 9, respectively. Examples of H^S multiplying ${\sigma }_c^2$ are shown in Supplementary Figure S9. Model selection was performed using Bayesian information criterion (BIC). Except for Height 2018, which is shown here for model C for direct comparison with the 2010 and 2014 GWASs, and whose model D fit is shown in Supplementary Figure S35, missing model D parameters indicate that those parameters could not reliably be estimated. For Crohn’s disease, ulcerative colitis, and total cholesterol $p_d(L)=p_{d1}$ for all L. Estimated BIC values for three models (B, C, and D) are shown in Appendix Table D1: the 3-parameter model B with only the “b” Gaussian (${\pi }_1,{\sigma }_b,{\sigma }_0$); the 8-parameter model C with both the “b” with “c” Gaussians (Equation 4); and the 12-parameter model D with “b,” “c,” and “d” Gaussians (Equation 5). Ninety-five percent confidence intervals are in Appendix Tables E1–E3.