Table 2 20 proteins with the greatest heterogeneity in inter-platform correlation estimates across contributing ancestry groups

From: Cross-ancestry comparison of aptamer and antibody protein measures

Protein summary

Probe association with ancestry-differentiated PAV

cis-pQTL

Correlation estimates (Pearson’s r) per ancestry and

cross-ancestry heterogeneity

UniProt

Target full name

SeqId

OlinkID

SomaScan

Olink

ALL

AFR

AMR

EAS

EUR

Q

Q p value

P05154

Plasma serine protease inhibitor

seq.3389.7

OID30763

TRUE

TRUE

−0.27

−0.38

−0.10

0.04

−0.28

39.41

1.42E-08

P12318

Low affinity immunoglobulin gamma Fc region receptor II-a

seq.3309.2

OID20391

FALSE

FALSE

0.30

0.18

0.38

−0.02

0.46

70.13

4.01E-15

Q96HD1

Cysteine-rich with EGF-like domain protein 1

seq.7628.40

OID30619

FALSE

FALSE

0.17

−0.01

0.18

0.25

0.26

25.10

1.47E-05

Q86SJ6

Desmoglein-4

seq.22568.7

OID21325

FALSE

FALSE

0.07

0.02

0.02

0.10

0.31

37.32

3.93E-08

Q9Y3E2

BolA-like protein 1

seq.15370.5

OID30837

FALSE

FALSE

0.28

0.37

0.13

0.49

0.22

31.39

7.03E-07

P17813

Endoglin

seq.4908.6

OID20287

FALSE

FALSE

0.39

0.40

0.38

0.16

0.47

26.78

6.54E-06

Q9H4A9

Dipeptidase 2

seq.8327.26

OID21305

FALSE

FALSE

0.31

0.46

0.34

0.17

0.19

32.83

3.50E-07

Q6UWN8

Serine protease inhibitor Kazal-type 6

seq.5731.1

OID21450

TRUE

FALSE

0.33

0.23

0.37

0.55

0.45

30.38

1.15E-06

Q03405

Urokinase plasminogen activator surface receptor

seq.2652.15

OID20764

TRUE

TRUE

0.44

0.24

0.61

0.64

0.47

55.64

5.02E-12

P41439

Folate receptor gamma

seq.15495.9

OID21485

FALSE

FALSE

0.47

0.61

0.24

0.43

0.42

34.70

1.41E-07

P15289

Arylsulfatase A

seq.3583.54

OID21138

FALSE

FALSE

0.40

0.47

0.49

0.43

0.25

28.56

2.76E-06

P02008

Hemoglobin subunit zeta

seq.6919.3

OID30307

FALSE

FALSE

0.40

0.52

0.29

0.50

0.29

31.88

5.56E-07

P04217

Alpha-1B-glycoprotein

seq.16561.9

OID30771

FALSE

FALSE

0.54

0.45

0.29

0.58

0.61

38.09

2.71E-08

Q13790

Apolipoprotein F

seq.12370.30

OID30701

FALSE

FALSE

0.49

0.29

0.48

0.67

0.55

50.38

6.62E-11

A6NHS7

MANSC domain-containing protein 4

seq.9578.263

OID30141

FALSE

FALSE

0.45

0.31

0.30

0.60

0.48

32.80

3.55E-07

P51858

Hepatoma-derived growth factor

seq.16758.96

OID21455

FALSE

FALSE

0.48

0.65

0.31

0.48

0.45

36.83

4.99E-08

Q9NQ38

Serine protease inhibitor Kazal-type 5

seq.8028.22

OID21148

FALSE

TRUE

0.49

0.32

0.56

0.62

0.51

29.97

1.40E-06

P36959

GMP reductase 1

seq.19254.125

OID20641

TRUE

FALSE

0.42

0.65

0.35

0.60

0.32

71.80

1.75E-15

O14791

Apolipoprotein L1

seq.11510.31

OID30708

TRUE

TRUE

0.54

0.33

0.54

0.58

0.55

27.54

4.54E-06

O15455

Toll-like receptor 3

seq.16918.198

OID20612

FALSE

FALSE

0.58

0.38

0.55

0.63

0.60

29.85

1.48E-06

  1. Displayed here are 20 out of the 80 proteins with Bonferroni-significant heterogeneity in inter-platform Pearson’s r correlation estimates across the 4 ancestry groups represented in this study. The 20 proteins displayed in this table show the greatest degree of heterogeneity in per-ancestry Pearson’s r correlations according to the lowest Cochran’s Q p values (generated in Cochran’s Q test of heterogeneity, one-sided test of significance), and the greatest range in Pearson’s r values across ancestry groups. “cis-pQTL summary” columns denote whether the SomaScan protein measure (SeqID) or the Olink protein measure (OlinkID) associates with an ancestry-differentiated protein altering variant (PAV) in cis- (i.e., cis-pQTL significantly associates with protein measure at traditional genome-wide significance (p < 5 × 10−8) and is driven by a PAV with ancestry-differentiated allele frequencies according to Chi-square test (one-sided test of significance) using allele frequencies derived in each contributing ancestry group, with significance defined as a X2 value in the 75th percentile). pQTL were generated with linear regression under an additive model (two-sided test of significance). Additional details about these proteins, as well as the remaining 60 proteins with Bonferroni-significant ancestry-heterogeneous inter-platform correlations can be found in Supplementary Data 3.
  2. AFR African, AMR Admixed American, EAS East Asian, EUR European Ancestry.
Back to article page