ECM-LSE: Prediction of Extracellular Matrix Proteins Using Deep Latent Space Encoding of k-Spaced Amino Acid Pairs

Ubaid M Al-Saggaf; Muhammad Usman; Imran Naseem; Muhammad Moinuddin; Ahmad A Jiman; Mohammed U Alsaggaf; Hitham K Alshoubaki; Shujaat Khan

doi:10.3389/fbioe.2021.752658

ECM-LSE: Prediction of Extracellular Matrix Proteins Using Deep Latent Space Encoding of k-Spaced Amino Acid Pairs

Front Bioeng Biotechnol. 2021 Oct 14:9:752658. doi: 10.3389/fbioe.2021.752658. eCollection 2021.

Authors

Ubaid M Al-Saggaf^{1

2}, Muhammad Usman³, Imran Naseem^{4

5

6}, Muhammad Moinuddin^{1

2}, Ahmad A Jiman², Mohammed U Alsaggaf^{1

7}, Hitham K Alshoubaki^{1

2}, Shujaat Khan⁸

Affiliations

¹ Center of Excellence in Intelligent Engineering Systems, King Abdulaziz University, Jeddah, Saudi Arabia.
² Electrical and Computer Engineering Department, King Abdulaziz University, Jeddah, Saudi Arabia.
³ Department of Computer Engineering, Chosun University, Gwangju, South Korea.
⁴ Research and Development, Love For Data, Karachi, Pakistan.
⁵ School of Electrical, Electronic and Computer Engineering, The University of Western Australia, Perth, WA, Australia.
⁶ College of Engineering, Karachi Institute of Economics and Technology, Korangi Creek, Karachi, Pakistan.
⁷ Department of Radiology, Faculty of Medicine, King Abdulaziz University, Jeddah, Saudi Arabia.
⁸ Department of Bio and Brain Engineering, Daejeon, South Korea.

Abstract

Extracelluar matrix (ECM) proteins create complex networks of macromolecules which fill-in the extracellular spaces of living tissues. They provide structural support and play an important role in maintaining cellular functions. Identification of ECM proteins can play a vital role in studying various types of diseases. Conventional wet lab-based methods are reliable; however, they are expensive and time consuming and are, therefore, not scalable. In this research, we propose a sequence-based novel machine learning approach for the prediction of ECM proteins. In the proposed method, composition of k-spaced amino acid pair (CKSAAP) features are encoded into a classifiable latent space (LS) with the help of deep latent space encoding (LSE). A comprehensive ablation analysis is conducted for performance evaluation of the proposed method. Results are compared with other state-of-the-art methods on the benchmark dataset, and the proposed ECM-LSE approach has shown to comprehensively outperform the contemporary methods.

Keywords: amino acid composition (AAC); auto-encoder; classification; composition of k-spaced amino acid pair (CKSAAP); extracellular matrix (ECM); latent space learning; neural network.