 In this paper, we investigated two types of biomedical relations, drug-drug and chemical-protein interactions. To identify these relations, we used various transfer transformers including GPT-3, T5-SLIM underscore deck, and BERTGAT. Our results showed that transformer-based approaches focused solely on word relationships are capable of understanding language well without additional knowledge such as structural information. Additionally, we found that incorporating syntactic structure can improve the performance of relation extraction.