Skip to main content

Table 3 Comparisons with Negatome and TissueNet by considering cancer types and normal counterparts (see Additional file 1: Table S5). The overlap is reported in percentage with respect to the number of total unique predictions per cancer/tissue. For each tissue, we report in the first row the percentage of INBIA’s overlapping, while in the second PERA’s one

From: INBIA: a boosting methodology for proteomic network inference

Cancer type

Negatome (%)

TissueNet (%)

BLCA

0.745

51.565

 

1.463

44.390

BRCA

0.247

47.654

 

0.830

29.461

COAD

0.609

44.901

 

0.806

34.274

  

30.323

GBM

0

36.452

  

28.064

  

25.484

  

19.636

 

0.727

26.909

  

20.727

  

15.636

HNSC

0.699

43.706

 

0.820

31.967

KIRC

1.608

48.231

 

2.137

40.171

LGG

0.62

32.812

 

1.020

25.510

LUAD

0.711

46.373

 

2.367

41.420

LUSC

0.895

46.418

 

3.297

43.407

OV

0.665

25.942

 

1.060

22.261

PRAD

0.602

39.458

 

0.395

32.411

READ

1.064

43.769

 

1.431

38.998

SKCM

0.446

54.018

 

0.704

42.958

STAD

0.315

43.218

 

0.772

35.135

THCA

0.308

26.769

  

40.308

 

0.763

19.466

  

26.336

UCEC

0.615

48.615

 

1.463

37.073

  1. Text highlighted in italic refers to our method (INBIA), the second one to PERA