Inferring Compound Similarity: A Clustering Approach in Drug Discovery

Drug discovery and development rely heavily on the systematic process of finding and altering chemical structures to achieve diverse biological activities for specific diseases. The primary goal of drug discovery is to find novel chemicals with medicinal promise. From a pool of millions of molecules...

Celý popis

Uloženo v:

Podrobná bibliografie
Vydáno v:	2024 1st International Conference on Cognitive, Green and Ubiquitous Computing (IC-CGU) s. 1 - 6
Hlavní autoři:	Palli, Parimala, Mishra, Satyasis, Rao, P. Srinivasa
Médium:	Konferenční příspěvek
Jazyk:	angličtina
Vydáno:	IEEE 01.03.2024
Témata:	Agglomerative Hierarchical Clustering Algorithm Clustering algorithms Drugs K-Means Clustering Algorithm Lead Machine Learning Machine learning algorithms SMILES Supervised Learning Systematics Task analysis Ubiquitous computing UnSupervised Learning
On-line přístup:	Získat plný text
Tagy:	Přidat tag Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!

Popis
Shrnutí:	Drug discovery and development rely heavily on the systematic process of finding and altering chemical structures to achieve diverse biological activities for specific diseases. The primary goal of drug discovery is to find novel chemicals with medicinal promise. From a pool of millions of molecules, drug designers must navigate convoluted and time-consuming development procedures in order to create a novel medicine. Finding the right lead chemical to optimize for advancement to preclinical trials and, eventually, an authorized medicine is the first step in target discovery, which is accomplished by mining a huge dataset of compounds. A more systematic and cost-effective approach to medication design and development is being made using deep learning and machine learning algorithms. The seven-step process of drug discovery includes pharmacogenic identification, target identification, lead optimization, preclinical testing, and clinical trials. Target identification and lead optimization are two areas that can benefit from machine learning techniques. Applying machine learning approaches to the task of finding a lead compound will greatly improve the efficiency of developing a novel medication to treat a condition. In comparison to the conventional drug development method, computational drug discovery is more efficient, leading to higher drug quality and lower process failure rates. This research proposes compound similarity prediction to the group compounds using clustering methods such as k-means, agglomerative hierarchical, and fuzzy c-means. Further, a similarity function based on the Simplified Molecular Inline Line Entry System (SMILES) is applied to determine the reaction of an unknown new substance by comparing its chemical structure to that of known chemicals. The drug development process can be accelerated with this aspect.
DOI:	10.1109/IC-CGU58078.2024.10530783