Table 1 Top 20 features used in the boosted trees of the AD-focused MPXgb model.

From: Machine learning prediction and tau-based screening identifies potential Alzheimer’s disease genes relevant to immunity

Feature

Data source

Gain

PPI:GAPDH

STRING

0.27852

PPI:IL10

STRING

0.08283

PPI:GSTP1

STRING

0.03823

Darunavir:HELA signature

LINCS

0.03732

Aminosalicylic acid:MCF7 signature

LINCS

0.03413

Blonanserin:PC3 signature

LINCS

0.03004

Amiloride:MC7 signature

LINCS

0.02594

Trifluridine:PC3 signature

LINCS

0.02048

Brain - Anterior cingulate cortex (BA24)

GTEx

0.01866

Travoprost:HA1E signature

LINCS

0.01848

Regulation of mRNA stability by proteins that bind AU-rich elements

Reactome

0.01775

PPI:JAK2

STRING

0.01729

PPI:BDNF

STRING

0.01684

PPI:IL2

STRING

0.01593

Levosulpiride:PC3 signature

LINCS

0.01556

Azelaic acid:MC7 signature

LINCS

0.01538

Saquinavir:A549 signature

LINCS

0.01534

Cyproterone:MC7 signature

LINCS

0.01411

Vorinostat:HT115 signature

LINCS

0.01138

Trifluoperazine:WSUDLCL2 signature

LINCS

0.01092

  1. The Gain highlights the relative contribution of each feature to the model.