APIN

APIN is a deep learning model to identify antimicrobial peptide (AMP) sequences. The model utilizes an embedding layer and a multi-scale convolutional network, which contains multiple convolutional layers with varying filter lengths. This architecture allows the model to capture and use latent features from the input sequences at different scales.

The authors also proposed a fusion model that incorporates additional information to further improve APIN's performance. The results demonstrate that both the base and fusion models outperform state-of-the-art models on two AMP datasets and the Antimicrobial Peptide Database (APD)3 benchmark dataset. Additionally, the fusion model surpasses the state-of-the-art model on an anti-inflammatory peptides (AIPs) dataset in terms of accuracy.