Binary AMR Prediction from PBP4 using Transformers
Benchmarking DNABERT and Nucleotide Transformer for antimicrobial resistance prediction, with parameter-efficient adaptation.
Benchmarked encoder-only pretrained transformer models, including DNABERT and Nucleotide Transformer, for binary antimicrobial resistance (AMR) prediction from pbp4 genomic sequences, and investigated representation learning and parameter-efficient adaptation of pretrained biological language models using PyTorch. Seminar “Deep Learning in Molecular Biology”, supervised by Prof. Dr. Alice C. McHardy.