Table 1 Summary of AiKPro dataset and strict split subsets.

From: AiKPro: deep learning model for kinome-wide bioactivity profiling using structure-based sequence alignments and molecular 3D conformer ensemble descriptors

Dataset name

Train/test

Data source

Kinases

Compounds

Bioactivity

AiKPro dataset

Train

BindingDB, DTC

391

156,284

337,171

Test

Metz

165

618

15,271

Strict split for docking

Train

BindingDB, DTC

391

156,284

337,171

Test

BindingDB, DTC, Metz

6

148

563

Strict split for kinases

Train

BindingDB, DTC

388

156,202

335,648

Test

Metz

3

1,505

1,522

Strict split for compounds

Train

BindingDB, DTC

391

155,781

287,493

Test

Metz

165

618

15,271

Back to article page