Skip to main content

Table 1 GAGE applied to the two lung cancer datasets of large sample sizes

From: GAGE: generally applicable gene set enrichment for pathway analysis

Boston study

Michigan study

Experimental Sets

p-val

q-val

Notes

Experimental Sets

p-val

q-val

Notes

Tarte_Plasma_Blastic

1.8E-64

1.1E-61

c

Tarte_Plasma_Blastic

5.6E-42

4.1E-39

c

Uvb_Nhek3_All

1.2E-59

3.6E-57

t

Cancer_Undifferentiat

1.0E-22

3.8E-20

bt

Peng_Glutamine_Dn

3.7E-59

7.6E-57

c

Brca_Er_Neg

8.3E-19

2.0E-16

bt

Lei_Myb_Regulated_G

5.8E-55

8.8E-53

bt, c

Serum_Fibroblast_Cell

3.2E-18

5.9E-16

bt, c

Peng_Leucine_Dn

4.0E-42

4.8E-40

c

Uvb_Nhek3_All

5.3E-17

7.7E-15

t

Cancer_Undifferentiat

3.0E-41

3.0E-39

bt

Caries_Pulp_Up

4.7E-16

4.6E-13

 

Brca_Er_Neg

2.0E-40

1.7E-38

bt

Zhan_Mm_Cd138_Pr_

8.3E-15

1.0E-12

bt

Peng_Rapamycin_Dn

3.5E-38

2.7E-36

c

Li_Fetal_Vs_Wt_Kidne

3.7E-14

3.8E-12

t

Rcc_Nl_Up

5.2E-36

3.5E-34

t

Dox_Resist_Gastric_Up

1.2E-13

1.1E-11

bt

Cancer_Neoplastic_Me

4.2E-35

2.6E-33

t

Idx_Tsa_Up_Cluster3

2.4E-13

1.9E-11

c

Canonical Pathways

p-val

q-val

Notes

Canonical Pathways

p-val

q-val

Notes

Gpcrs_Class_A_Rhod

9.2E-23

3.1E-20

bt

Gpcrs_Class_A_Rhod

3.1E-10

1.0E-07

bt

Gpcrdb_Class_A_Rho

4.7E-21

7.8E-19

bt

Gpcrdb_Class_A_Rho

1.1E-09

1.9E-07

bt

Blood_Clotting_Casca

5.1E-15

4.7E-13

bt

Androgen_Genes

5.2E-08

5.8E-06

bt

Intrinsicpathway

6.3E-15

5.3E-13

bt

Cytokinepathway

1.9E-07

1.6E-05

bt

Fibrinolysispathway

1.1E-12

9.1E-11

bt

Prostaglandin_And_Le

2.9E-05

2.4E-03

bt

Peptide_Gpcrs

1.9E-12

1.6E-10

bt

Proliferation_Genes

5.1E-05

4.3E-03

c

Tyrosine_Metabolism

8.7E-09

7.3E-07

bt

Peptide_Gpcrs

5.8E-05

4.8E-03

bt

Extrinsicpathway

5.5E-07

4.6E-05

bt

Intrinsicpathway

9.1E-05

7.6E-03

bt

Gpcrdb_Other

5.2E-06

4.4E-04

bt

Androgen_And_Estrog

4.2E-04

3.4E-02

bt

Small_Ligand_Gpcrs

6.7E-06

5.6E-04

bt

Blood_Clotting_Casca

7.5E-04

5.9E-02

bt

  1. Top 10 most significantly enriched experimental sets and canonical pathways in poor clinical outcomes vs good outcomes were inferred by GAGE from two published lung adenocarcinoma data sets used in the GSEA paper [3]. Both positively and negatively regulated gene sets were collected and ranking by the p-value, and by absolute value of average t-statistics (data not shown) for ties. FDR q-values were estimated to correct the p-values for the multiple testing issue. Consistencies between the two data sets are shown in bold font. Notes show the connections of the gene sets to cancer related topics: t for tumor related; bt for tumor metastasis and bad outcome; c for cell growth and proliferation related; and blank represents no evident connection. These annotations came from the original studies for experimental sets, or made based on more than three independent literature works for the canonical pathway.