®®®® SIIA Público

Título del libro: Applications In Software Engineering - Proceedings Of The 13th International Conference On Software Process Improvement, Cimps 2024
Título del capítulo: Network Based Machine Learning Description of the Antidepressive ADR Landscape

Autores UNAM:
GUILLERMO DE ANDA JAUREGUI; ENRIQUE HERNANDEZ LEMUS;
Autores externos:

Idioma:

Año de publicación:
2024
Palabras clave:

Adverse drug reactions; Antidepressant; Graph embeddings; Graph2vec; Machine-learning; Mental disorders; Network-based; Random forest regression; Random forests; Xgboost regression


Resumen:

© 2024 IEEE.Depression is a common mental disorder that affects the lives of millions of people around the world. Antidepressants are used to treat this disorder, whoever they lead to adverse drug reactions (ADRs). Adverse drug reactions (ADRs) are harmful or unpleasant reactions produced by an intervention when using pharmacological substances. Studies have correlated the nonadherence of antidepressant treatments with ADRs. It is important to predict potential ADRs and create therapies that reduce ADRs to improve adherence from the patient. Here, we applied methods -using knowledge discovery in the Food and Drug Administration (FDA) Adverse Event Reporting System (FAERS) database, graph embeddings of the chemical structure of the drugs, and machine learning methods (random forest and XGBoost) - to predict potential adverse drug reactions. We analyzed a set of 28 antidepressant drugs and identified two main features that were significant in predicting ADRs when comparing with other antidepressants.


Entidades citadas de la UNAM: