Table 1 Success rate of different modelling setups.

From: Improved prediction of protein-protein interactions using AlphaFold2

Neural network configuration

NN model

model_1

model_1

model_1_ptm

model_1_ptm

Recycles

10

3

10

3

Ensembles

1

8

1

8

Setup short name

m1-10-1

m1-3-8

mp-10-1

mp-3-8

Paired MSA

28.7

28.2

28.7

27.8

AF2 MSA

31.5

33.3

26.4

23.6

AF2+Paired MSA

39.4

38.4

32.4

31.0

  1. Results of AF2 run on the development set (n = 216) using different MSAs and neural network configurations. Row labels in bold indicate AF2 setup features. Every column in the table refers to an overall setup and every corresponding value refers to a run of the described setup with a different input MSA. Values represent the percentage of acceptable models (DockQ ≥ 0.23) overall the development set. The highest success rates for each MSA type are highlighted in bold.
Back to article page