A carregar...
Publicação
A inteligência artificial na descoberta de novos medicamentos
| Nome: | Descrição: | Tamanho: | Formato: | |
|---|---|---|---|---|
| 1.12 MB | Adobe PDF |
Autores
Orientador(es)
Resumo(s)
A descoberta e desenvolvimento de medicamentos é uma área complexa, que engloba várias etapas, levando a um processo bastante moroso e dispendioso. Com o aumento da digitalização de dados, a inteligência artificial impulsionou a sua aplicação em diversos setores da sociedade, sendo que o setor farmacêutico aproveitou e implementou o seu uso neste processo.
Melhorias notáveis no poder computacional combinadas com desenvolvimentos em tecnologia artificial podem ser usadas para transformar o processo de desenvolvimento de medicamentos, uma vez que podem ser aplicadas em todas as etapas desse processo – descoberta de medicamentos, desenvolvimento pré-clínico, desenvolvimento clínico e pós- comercialização – incluindo várias aplicações em propriedades ou previsões de atividade como propriedades físico-químicas, absorção, distribuição, metabolismo, excreção e toxicidade ou relações quantitativas estrutura-propriedade (QSPR) ou relações quantitativas estrutura- atividade (QSAR).
Recentemente, o design de medicamentos entrou na era do big data e os métodos de machine learning evoluíram gradualmente para métodos de deep learning com processamento de big data mais robusto e eficaz, levando à combinação de inteligência artificial e tecnologia de design de medicamentos assistida por computador.
O design de novo ajuda no processo de descoberta de fármacos ao criar novos agentes farmacêuticos ativos com propriedades pretendidas de forma económica e tempo-eficiente.
O principal benefício da inteligência artificial é a diminuição do tempo necessário para o desenvolvimento de medicamentos e, portanto, os custos associados ao processo, melhorando o retorno do investimento, pode levar a uma redução de custos para o consumidor final.
Uma vez que um tamanho considerável de training data é necessário para que o training em deep learning seja um sucesso e a acessibilidade a estes dados por vezes não seja adequada para que a inteligência artificial seja eficaz, ainda há muito espaço para melhorias em termos de precisão do método, apesar do seu sucesso crescente.
The discovery and development of drug products is a complex area, which involves numerous stages, leading to a very time consuming and expensive process. With the increase of data digitization, artificial intelligence has recently boosted its application in various sectors of society, and the pharmaceutical sector has taken advantage of and implemented its use in this process. Notable improvements in computational power combined with developments in artificial technology could be used to transform the drug development process since they can be applied to support all steps of this process – drug discovery, preclinical development, clinical development and postmarketing, – including various applications in property or activity predictions like physicochemical properties, absorption, distribution, metabolism, excretion and toxicity or quantitative structure-property relationships (QSPR) or quantitative structure- activity relationships (QSAR). Recently, drug design has entered the era of big data and machine learning methods have gradually evolved into deep learning methods with stronger and more effective big data processing, leading to the combination of artificial intelligence and computer-assisted drug design technology. The de novo design helps drug discovery projects by creating novel pharmaceutically active agents with desired properties in a cost- and time-efficient manner. The main benefit of artificial intelligence is that it decreases the time that is needed for drug development and therefore the costs associated with the process, improves the return on investment and may even lead to a cost reduction for the end user. Since a considerable size of training data is required for deep learning training to be a success and accessibility to said data is not adequate at times for artificial intelligence to be effective, there is still a big room for improvement in terms of method accuracy, despite its rising success.
The discovery and development of drug products is a complex area, which involves numerous stages, leading to a very time consuming and expensive process. With the increase of data digitization, artificial intelligence has recently boosted its application in various sectors of society, and the pharmaceutical sector has taken advantage of and implemented its use in this process. Notable improvements in computational power combined with developments in artificial technology could be used to transform the drug development process since they can be applied to support all steps of this process – drug discovery, preclinical development, clinical development and postmarketing, – including various applications in property or activity predictions like physicochemical properties, absorption, distribution, metabolism, excretion and toxicity or quantitative structure-property relationships (QSPR) or quantitative structure- activity relationships (QSAR). Recently, drug design has entered the era of big data and machine learning methods have gradually evolved into deep learning methods with stronger and more effective big data processing, leading to the combination of artificial intelligence and computer-assisted drug design technology. The de novo design helps drug discovery projects by creating novel pharmaceutically active agents with desired properties in a cost- and time-efficient manner. The main benefit of artificial intelligence is that it decreases the time that is needed for drug development and therefore the costs associated with the process, improves the return on investment and may even lead to a cost reduction for the end user. Since a considerable size of training data is required for deep learning training to be a success and accessibility to said data is not adequate at times for artificial intelligence to be effective, there is still a big room for improvement in terms of method accuracy, despite its rising success.
Descrição
Trabalho Final de Mestrado Integrado, Ciências Farmacêuticas, 2022, Universidade de Lisboa, Faculdade de Farmácia.
Palavras-chave
Inteligência artificial Machine learning Deep learning Descoberta e desenvolvimento de fármacos Mestrado integrado - 2022