Journal
ARTIFICIAL INTELLIGENCE IN MEDICINE
Volume 131, Issue -, Pages -Publisher
ELSEVIER
DOI: 10.1016/j.artmed.2022.102349
Keywords
Deep neural network; Anticancer peptides; Word embedding; FastText; Classification
Funding
- Mohammad Ibn Saud Islamic University [RG-21-51-01]
Ask authors/readers for more resources
This paper presents a FastText-based word embedding strategy and a deep neural network model for predicting anticancer peptides, achieving high accuracy and outperforming existing predictors. The proposed cACP-DeepGram model could be a reliable tool for scientists in academic research and drug discovery.
Cancer is a Toxic health concern worldwide, it happens when cellular modifications cause the irregular growth and division of human cells. Several traditional approaches such as therapies and wet laboratory-based methods have been applied to treat cancer cells. However, these methods are considered less effective due to their high cost and diverse side effects. According to recent advancements, peptide-based therapies have attracted the attention of scientists because of their high selectivity. Peptide therapy can efficiently treat the targeted cells, without affecting the normal cells. Due to the rapid increase of peptide sequences, an accurate prediction model has become a challenging task. Keeping the significance of anticancer peptides (ACPs) in cancer treatment, an intelligent and reliable prediction model is highly indispensable. In this paper, a FastText-based word embedding strategy has been employed to represent each peptide sample via a skip-gram model. After extracting the peptide embedding descriptors, the deep neural network (DNN) model was applied to accurately discriminate the ACPs. The optimized parameters of DNN achieved an accuracy of 96.94 %, 93.41 %, and 94.02 % using training, alternate, and independent samples, respectively. It was observed that our proposed cACP-DeepGram model outperformed and reported similar to 10 % highest prediction accuracy than existing predictors. It is suggested that the cACP-DeepGram model will be a reliable tool for scientists and might play a valuable role in academic research and drug discovery. The source code and the datasets are publicly available https://gitub.com/shahidak-barcs/cACP-DeepGram.
Authors
I am an author on this paper
Click your name to claim this paper and add it to your profile.
Reviews
Recommended
No Data Available