In all areas of tropical medicine and surgery, including emerging infectious diseases, neglected tropical diseases, non-communicable diseases, tropical biomedicine, tropical diseases, tropical fish diseases, tropical fish medicine, tropical health nutrition, tropical medicine, tropical medicine and health, and tropical medicine and hygiene, significant and influential scientific works are featured in the open access journal Tropical Medicine & Surgery. The journal's focus extends beyond merely tropical medicine; Equally welcome are publications on parasitology, veterinary medicine, and epidemiology that are closely related to tropical medicine research. It is a blessing for researchers and students who want to stay up to date on the most recent developments in the medical area because it will enable them to expand on their present knowledge.

PV has shown to be effective at identifying post-marketing adverse medication occurrences (ADE). Natural language processing (NLP) software was previously employed in study to extract unstructured texts pertinent to ADEs. However, sentences devoid of context make these algorithms less effective. Our goal was to create and test aTarantula, an unique NLP tool that uses an aggregated lexicon and a context-aware machine learning algorithm to identify pre-existing ADEs in social media.

In order to extract contextual data from three warfarin-using patient forums (MedHelp, MedsChat, and PatientInfo), aTarantula used FastText embeddings and an aggregated lexicon. From the UMLS and FAERS databases, the lexicon used warfarin package inserts and synonyms of warfarin ADEs. Three clinical pharmacists manually refined and validated the data once it had been placed on SQLite.

The most common organ systems for ADE reports were multiple organ systems (1.50%), followed by CNS side effects (1.19%). The least frequent side effect, lymphatic system ADEs, was observed in 0.09% of cases. Between patient-reported information from the forums and FAERS, the overall Spearman rank correlation coefficient was 0.19. The sensitivity and specificity of Tarantula were shown to be 84.2% and 98%, respectively, by pharmacist validation. Our findings were manually confirmed by three clinical pharmacists. Finally, we developed a lexicon that has been combined for social media ADE mining.

We were able to use aTarantula, a machine-learning algorithm based on artificial intelligence, to automatically extract ADEs relevant to warfarin from online social discussion forums. Our research demonstrates that a Tarantula can be used to find ADEs. On the varied dataset, future researchers can validate aTarantula.