Please use this identifier to cite or link to this item:
http://repository.ipb.ac.id/handle/123456789/165124| Title: | Penapisan Senyawa Herbal sebagai Agen Antipenuaan menggunakan Algoritma Random Forest |
| Other Titles: | Screening Herbal Compounds as Antiaging Agents using Random Forest Algorithm |
| Authors: | Kusuma, Wisnu Ananta Sitanggang, Imas Sukaesih ARIYANI, TAN, MARIA PUTRI |
| Issue Date: | 2025 |
| Publisher: | IPB University |
| Abstract: | Proses penuaan merupakan proses biologis alami yang ditandai dengan penurunan fungsi fisiologis secara bertahap serta peningkatan risiko penyakit terkait usia. Pada era modern, kesadaran masyarakat akan pentingnya memperhatikan tanda-tanda penuaan untuk menjaga kesehatan dan kualitas hidup, semakin meningkat. Hal ini mendorong upaya pengembangan bahan aktif yang tidak hanya efektif tetapi juga aman, seperti senyawa herbal. Penelitian ini bertujuan untuk mengidentifikasi senyawa herbal yang memiliki potensi sebagai agen antipenuaan melalui analisis interaksi antara senyawa dan target protein menggunakan algoritma Random Forest. Tahapan penelitian meliputi akuisisi data, praproses data, pembagian data, pemodelan, evaluasi model, dan prediksi senyawa herbal. Data protein terkait antipenuaan diperoleh dari basis data OMIM, sedangkan data senyawa diambil dari basis data BindingDB, PubChem, dan IJAH Analytics. Algoritma Random Forest dioptimasi dengan proses hyperparameter tuning untuk menghasilkan model yang optimal. Evaluasi model dilakukan menggunakan metrik
accuracy, precision, recall, F1-score, dan AUROC. Hasil pemodelan terbaik diperoleh dengan kombinasi deskriptor AAC dan metode penyeimbangan SMOTEENN, menghasilkan accuracy sebesar 96,74%, precision 97,36%, recall 97,27%, F1-score 97,32%, dan AUROC 96,59%. Model optimal ini berhasil memprediksi 130 senyawa herbal untuk antipenuaan. Aging is a natural biological process characterized by a progressive decline in physiological functions and an increased risk of age-related diseases. In the modern era, public awareness of the importance of delaying aging signs to maintain health and quality of life has grown significantly. This has driven the search for active ingredients that are not only effective but also safe, such as herbal compounds. This study aims to identify potential herbal anti-aging agents through interaction analysis between compounds and target proteins using the Random Forest algorithm. The research workflow consisted of data acquisition, preprocessing, splitting, modeling, model evaluation, and compound prediction. Protein targets related to aging were obtained from the OMIM database, while compound data were sourced from BindingDB, PubChem, and IJAH Analytics. The Random Forest algorithm was optimized through hyperparameter tuning. Model performance was evaluated using metrics including accuracy, precision, recall, F1-score, and AUROC. The best performance was achieved using a combination of AAC descriptors and the SMOTEENN balancing method, yielding an accuracy of 96,74%, precision of 97,36%, recall of 97,27%, F1-score of 97,32%, and AUROC of 96,59%. Using this optimal model, 130 herbal compounds were successfully predicted to have anti-aging potential. |
| URI: | http://repository.ipb.ac.id/handle/123456789/165124 |
| Appears in Collections: | UT - Computer Science |
Files in This Item:
| File | Description | Size | Format | |
|---|---|---|---|---|
| cover_G6401211049_9f966177e8914e0989503ca2aa077aa0.pdf | Cover | 514.73 kB | Adobe PDF | View/Open |
| fulltext_G6401211049_20c60f7a915641e8b190617b2ff5d3b0.pdf Restricted Access | Fulltext | 1.21 MB | Adobe PDF | View/Open |
| lampiran_G6401211049_d82cb7c3c97c4a39b24b2aebf7702a85.pdf Restricted Access | Lampiran | 354.47 kB | Adobe PDF | View/Open |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.