(For USM Staff/Student Only)

EngLib USM > Ω School of Electrical & Electronic Engineering >

Prediction of antimicrobial peptides based on sequence alignment and secondary structure sequence and segment sequence

Prediction of antimicrobial peptides based on sequence alignment and secondary structure sequence and segment sequence / Soh Meng Wah
Peptida antimicrobial (AMP) adalah sejenis peptide semula jadi yang penting untuk sistem imun. Penyelidik berminat untuk membuat ubat dengan AMP sebagai alternatif kerana bakteria semakin boleh menentang dengan antibiotik yang sedia ada. Walaubagaimanapun, eksperimen untuk mengekstrak AMP dari protein mahal dan mengambil masa. Oleh itu, alat pengiraan yang berkesan dan tepat meramalkan AMP baru amat dikehendaki untuk mengkaji ubat baru. Dalam projek ini, algoritma baru dicadangkan sebagai alat pengiraan dengan mengabungkan kaedah penjajaran urutan dan urutan struktur sekunder (SSS) dan urutan segmen (SS). Penjajaran urutan dilaksana berdasarkan HSPs maksimum skor yang diramalkan oleh BLASTP. Kaedah penjajaran urutan tidak dapat meramalkan semua urutan. Keputusan fasa penjajaran urutan adalah di 91.02 % bagi set data biasa, 80.88 % untuk urutan yang mempunyai persamaan <0.7, dan 96.02 % untuk CAMP set data. Bagi urutan yang tidak boleh diramalkan, ramalan diteruskan dengan menggunakan ciri-ciri SSS dan SS. Pengekstrakan ciri dan pilihan ciri dilakukan dan kemudian ciri-ciri tersebut digunakan untuk melatih pembelajaran mesin SVM bagi mengklasifikasikan urutan sama ada AMP atau bukan AMP. Keputusan ujian keseluruhan adalah 83.27% bagi set data biasa, 71.83% untuk urutan yang mempunyai persamaan <0.7, dan 91.49% untuk CAMP set data. Berbanding dengan fasa kedua kajian dulu yang menggabungkan dengan kaedah penjajaran jujukan, kajian ini mempunyai hasil yang rendah (<27%) dengan hanya menggunakan ramalan dengan SSS dan SS. Ini menunjukkan bahawa algoritma baru yang dicadangkan tidak sesuai untuk digunakan sebagai peramal AMP. ___________________________________________________________________________________ Antimicrobial peptides (AMPs) are natural peptides that are important for immune system. Researchers are interested in designing alternative drugs with AMPs because more bacteria are becoming resistant to the available antibiotics. However, the experiments to extract AMP from protein sequences are time consuming and costly. Thus, a computational tool with more effective and accurately predicting novel AMPs is highly demanded to provide more candidates and useful insights for drug design. In this study, a new algorithm is proposed as a computational tool by integrating the sequence alignment method and the secondary structure sequence (SSS) and segment sequence (SS). The sequence alignment is accomplished by the classification of test sequences based on the maximum high-scoring segment pairs (HSPs) score predicted by Basic Local Alignment Search Tool for protein (BLASTP). The results of sequence alignment phase are in 91.02% for normal dataset, 80.88% on <0.7 sequence similarity train set and 96.02% for CAMP dataset. Sequence alignment method is not able to predict all sequences and the unpredicted sequences is then predicted by utilizing the SSS and SS features. Feature extraction and feature selection is performed to obtain the features. These features are used to train the SVM model which is then be used to classify the sequences to whether it is AMP or non-AMP. The overall results of independent test is 83.27% for normal dataset, 71.83% for sequence with <0.7 similarity dataset and 91.49% for CAMP dataset. In comparison of second phase with past research that combines with sequence alignment method, this research has relatively low yield (<27%) contributed by the prediction utilizing SSS and SS features only. This indicates that the proposed algorithm is not suitable to be used as AMPs predictor.
Contributor(s):
Soh Meng Wah - Author
Primary Item Type:
Thesis
Identifiers:
Accession Number : 875005917
Language:
English
Subject Keywords:
Antimicrobial peptides (AMPs); natural peptides; immune system
First presented to the public:
8/1/2015
Original Publication Date:
6/5/2018
Previously Published By:
Universiti Sains Malaysia
Place Of Publication:
School of Electrical & Electronic Engineering
Citation:
Extents:
Number of Pages - 88
License Grantor / Date Granted:
  / ( View License )
Date Deposited
2018-06-05 10:23:08.671
Date Last Updated
2019-01-07 11:24:32.9118
Submitter:
Mohd Jasnizam Mohd Salleh

All Versions

Thumbnail Name Version Created Date
Prediction of antimicrobial peptides based on sequence alignment and secondary structure sequence and segment sequence1 2018-06-05 10:23:08.671