Browsing by Author "ANUM RAUF"

Now showing 1 - 1 of 1
  • No Thumbnail Available
    Item
    Boosted deep convolutional neural network based antihypertensive peptide predictor
    (UMT.Lahore, 2019) ANUM RAUF
    Heart attack and other heart-related diseases are the main cause of fatalities in the world. These diseases and some other severe diseases like kidney failure and disability are mainly caused by hypertension. There are several drugs available for hypertension treatment but they sometimes cause severe side effects as well. Bioactive peptides that are derived from natural sources and have antihypertensive activity in them can function as likely replacements to pharmacological drugs with no or very fewer side effects. Although these peptides are beneficial it is costly, time-consuming, and need lots of experimentation to find these drugs. So, to overcome this problem we have proposed an automated antihypertensive peptides prediction system which is able to predict whether the given peptide is antihypertensive or not. Recently few researches have conducted for this as well and have satisfactory results but there is a need for more improvement. In this research, we have used two peptide’s sequence datasets (benchmarking and independent dataset) and applied two feature extraction (Amino acid composition and Dipeptide composition) techniques on them. Through these techniques, we have converted the peptide’s sequence dataset into image dataset and further applied deep learning algorithm (convolutional neural network) and machine learning algorithm (support vector machine) on them. While comparing our model with existing models, our model showed greater results with the approximate enhancement of 7-8% of accuracy for both datasets.