Benchmarking Large Language Models on Multiple Tasks in Bioinformatics NLP with Prompting