Prediction of Ubiquitination Sites Using UbiNets

Table 2

Classification models and their description.

Models	Model description
ub_v1	Is comprised of a single UbiNet Block
ub_deeper	A deeper network, is comprised of 2 UbiNet Blocks
ub_deeper_x2	The deepest UbiNet architecture, is comprised of 3 UbiNet Blocks
Dense_residual_v2	A shallow residual network inspired design
50x2_mlp_elu	2 hidden layer feed forward network with ELU activation and 50 nodes per hidden layer
100x2_mlp_elu	2 hidden layer feed forward network with ELU activation and 100 nodes per hidden layer
50x2_mlp_relu	2 hidden layer feed forward network with ReLU activation and 50 nodes per hidden layer
100x2_mlp_relu	2 hidden layer feed forward network with ReLU activation and 100 nodes per hidden layer
Random forest, 100 estimators	Random forest of 100 metaestimators
Random forest, 200 estimators	Random forest of 200 metaestimators
gbm, 100 estimators	Gradient Boosting Machine is comprised of 100 metaestimators