Deep learning–based language models, such as BERT, T5, XLNet and GPT, are promising for analyzing speech and texts. In recent years, however, they have also been applied in the fields of biomedicine and biotechnology to study genetic codes and proteins.