Implementation of Enzyme Family Classification by using Autoencoders in a Study Case with Imbalanced and Underrepresented Classes

Figure 1.

Figure 2.

Figure 3.

Figure 4.

Figure 5.

Figure 6.

Results of the family classification

	Precision	Recall	F1-Score
GH18	0.90	0.90	0.90
GH19	0.91	1.00	0.95
No Enzyme	0.89	0.80	0.84
accuracy	0.90
macro avg	0.90	0.90	0.90
weighted avg	0.90	0.90	0.90

Number of enzymes per family

Family	Number of enzymes
GH18	356
GH19	83

Comparison of the trainings of the First Level

	Loss Function	Loss Function (Validation)
Without SMOTE	0.0202	0.0223
SMOTE	0.0127	0.0162
Hyperparameter optimization. (SMOTE)	0.0025	0.0074

Comparison of the training of the Second Level

	Loss Function	Loss Function (Validation)
Without SMOTE	0.0328	0.1163
SMOTE	0.0518	0.0330
Hyperparameter optimization. (SMOTE)	0.0392	0.0350

Comparison of different softwares for the classification of sequences into enzymes or non-enzymes (precision)

	EzyPred	ECPred	Proteinfer	AE
Not Enzyme	0.59	0.57	0.47	0.91
Enzyme	1.00	0.82	0.95	1.00

Comparison of different softwares for the classification of sequences into enzymes or non-enzymes (F1-score)

	EzyPred	ECPred	Proteinfer	AE
Not Enzyme	0.74	0.47	0.62	0.95
Enzyme	0.87	0.86	0.78	0.98

Comparison of different softwares for the classification of sequences into enzymes or non-enzymes (recall)

	EzyPred	ECPred	Proteinfer	AE
Not Enzyme	1.00	0.40	0.90	1.00
Enzyme	0.77	0.90	0.67	0.97