TNFR-LSTM: A Deep Intelligent Model for Identification of Tumour Necroses Factor Receptor (TNFR) Activity

Area of Science:

• Biochemistry
• Computational Biology
• Immunology

Background:

• Tumour necrosis factors (TNFs) are crucial in inflammation, cancer, and autoimmune diseases.
• Accurate identification of TNFs is challenging due to complex cytokine interactions.
• Existing machine learning models have limitations in reliably distinguishing TNFs.

Purpose of the Study:

• To develop an advanced model, DEEP-TNFR, for predicting tumour necrosis factor receptor (TNFR) activity.
• To enhance the accuracy of TNFR identification using deep learning.

Main Methods:

• Developed DEEP-TNFR model incorporating relative/reverse positions and statistical moments.
• Explored six deep learning classifiers: FCN, CNN, RNN, LSTM, Bi-LSTM, GRU.
• Evaluated model performance using self-consistency, independent set testing, and cross-validation (5- and 10-fold).

Main Results:

• LSTM classifier demonstrated superior performance among the tested deep learning models.
• DEEP-TNFR achieved high accuracy, specificity, sensitivity, and Matthews correlation coefficient.
• The model sets a new benchmark for TNFR identification compared to previous studies.

Conclusions:

• DEEP-TNFR significantly enhances the accuracy of TNFR identification.
• The model is expected to greatly aid ongoing research in inflammation, cancer, and autoimmune diseases.
• This work provides a robust computational tool for studying TNFRs.