Skip to main content

Table 2 Key drivers for cholesterol-associated gene subnetworks

From: Mergeomics: multidimensional data integration to identify pathogenic perturbations to biological systems

Subnetworks

−log10 P

Functional annotation

Top adipose KDs

Top liver KDs

Key driver

−log10 P

Co-hubs

Key driver

−log10 P

Co-hubs

Subnetwork 1 Lipoprotein

16.0

Lipid transport; cholesterol metabolism; lipoprotein; blood plasma

-

-

-

SPRY4

9.5

ABCG8

S100A10

4.5

-

Subnetwork 2 Lipid metabolism

8.1

Lipid metabolism; metalloprotein; oxidoreductase; endoplasmic reticulum

ACADVL

33.7

PPARA, CIDEA

FASN

49.0

GPAM, ACLY

FASN

26.8

ME1, ACSS2, ACLY, ELOVL6

SQLE

37.4

FDFT1, IDI1, MSMO1, NSDHL, HMGCS1, ALDOC

SCD

24.0

DNMT3L

DHCR7

26.9

PMVK, MUM1, FDPS, LSS, RDH11, MVD

CCBL2

23.3

-

HSD17B7

23.9

-

ACO2

23.0

AV075202, GPD2, NDUFV1

MMT00007490

18.8

HMGCR, LSS, FDFT1, MVD, ACSL3

Subnetwork 3 Immunoglobulin

6.1

Immunoglobulin V-set

COL1A1

12.4

-

COL6A3

21.4

-

COL1A2

9.4

COL3A1,COL2A1, MFAP2

VIM

11.0

-

OLFML3

8.8

-

CCDC3

10.4

OLFML3

POSTN

8.3

COL2A1

CXCR7

9.9

-

FN1

7.2

-

FBLN2

9.0

-

Subnetwork 4 ABC transport

5.0

ATP-binding cassette genes

-

-

-

SPRY4

12.0

ABCG8

MMT00062095

4.3

-

S100A10

3.2

-

Subnetwork 5 Retinoid metabolism

4.5

Retinoid metabolism; Visual transduction

-

-

-

GC

11.2

RBP4,APOH

TFPI2

3.2

-

AQP8

2.9

-

Subnetwork 6 Transcription

3.8

Transcription regulation; fatty acid metabolism; acyltransferase

SLC2A5

18.2

-

PKLR

23.6

MMT00060232, ELOVL6

ACADVL

17.7

PPARA, CIDEA

PNPLA5

19.0

ACLY, ACACA, PNPLA3

CPT2

15.9

-

PGD

12.2

-

EHHADH

15.1

-

FASN

11.6

GPAM,ACLY

ACO2

13.7

AV075202, GPD2, NDUFV1

INSIG1

10.7

-

  1. Initially, canonical pathways were evaluated for the enrichment of genetic perturbations to circulating cholesterol. As these pathways overlap with each other, non-redundant “subnetworks” were constructed that represent the most shared core genes between overlapping pathways. To verify the association with cholester, enrichment was re-evaluated for the subnetworks (second column in the table). Statistical significance was estimated as described in Table 1. Functional annotations were determined with the DAVID Bioinformatics Tool [45]. Key drivers and co-hubs were determined with the wKDA module within Mergeomics. Bayesian networks from multiple mouse studies were combined to create weighted adipose and liver consensus networks [43, 44]. Gene symbols were translated to human when available