THE PHILOSOPHY AND PROSPECTS OF FRAGMENT CONTRIBUTION ESTIMATIONS IN DRUG DISCOVERY

Shulga, D.; Ivanov, N.; Palyulin, V.

Авторы: Shulga D., Ivanov N., Palyulin V.
Сборник: Proceedings book of the XXVIII Symposium on Bioinformatics and Computer-Aided Drug Discovery
Год издания: 2022
Место издания: Institute of Biomedical Chemistry Moscow
Первая страница: 55
Последняя страница: 55
DOI: 10.18097/BCADD2022
Аннотация: Contemporary drug discovery faces hard times caused by stronger regulatory requirements andexhausting of "the low hanging fruits" of blockbusters’ targets. The systemic response of thepharmaceutical industry is the broadening of the Open Innovation style wide net of cooperation andvivid adoption of Agile approaches, that encourage fast and lean hypothesis testing during the drugdesign process. Definitely relevant in silico means which help to achieve the above goals arenecessary.Fragment in molecule is a very natural and still vague and disputable conception, which despitethe lack of strict scientific definition has been widely used in everyday practice. However this use hasbeen mostly restricted to qualitative applications. The focus of our study is to investigate the scope inwhich the decomposition of affinity in terms of separate contributions of different ligand fragmentsprovides useful and lean means, which can boost drug discovery process.Conceptually we derive our approach from the same principles that were laid into foundation ofthe well-known and appreciated fragment-based drug discovery (FBDD) approach. Moreover, weconsider our approach as a reverse version of FBDD conception, so we called it reverse fragment-based drug discovery (R-FBDD) [1]. However we have found that the perception of the twoapproaches, despite stemming from the same root, differs significantly, thus forming a phenomenon,which we call fragment-based perception hysteresis. This phenomenon, as well as its both true andfalse pros and cons will be presented.In the second part the recently proposed practical implementation of in silico fragmentcontribution estimation using scoring functions [1, 2] is employed to exemplify the potential practicaluse cases, where the proposed approach provides the researchers with additional valuable informationand guidance. For instance, our method can be used to rationally decide on which substituents aresuperfluous in a moderate activity hit compound, this use case being very common in lean practice. Inthis regard the group efficiency (GE) metrics for a certain group, which is a direct analog of the ligandefficiency (LE) for the entire ligand, can be used to guide decision to get rid of effectively ballastsubstituents. This move opens the opportunity of introduction of more effective substituents and thusleading not only to more active/affine ligand, but also to a structure with better ligand efficiency (LE).Other useful examples and use cases will be also presented.The proposed approach introduces additional dimensions, which can be used to rank differentsubstituents of the core structure. Moreover the value of the central scaffold can be critically evaluatedin order to make better judgments on how much further development using this scaffold is risky.The performed analysis reveals that the in silico fragment contribution approach is a promisingnew tool, which has deep thinking tradition behind it. Thus it is natural for researchers both to analyzeand to design in terms of fragments and their properties. An important property of our proposedapproach is its ease of application, which renders it as a valuable tool for lean and agile modern drugdiscovery, which heavily and wisely relies on in silico methods to boost the entire drugdiscovery process.
Добавил в систему: Шульга Дмитрий Александрович

	ИСТИНА	Войти в систему Регистрация
	ИПМех РАН
	Главная Поиск Статистика О проекте Помощь

ИСТИНА

ИПМех РАН

THE PHILOSOPHY AND PROSPECTS OF FRAGMENT CONTRIBUTION ESTIMATIONS IN DRUG DISCOVERYстатья Тезисы

Прикрепленные файлы