One unique trait of membrane proteins is that a significant fraction of their hydrophobic amino acids is exposed to the hydrophobic core of lipid bilayers rather than being embedded in the protein interior, which is often not explicitly considered in studies of membrane proteins. In our previous work, we proposed a characteristic and predictive quantity, the membrane contact probability (MCP), to describe the likelihood of the amino acids of a given sequence being in direct contact with the acyl chains of lipid molecules. We also developed a machine learning-based method for predicting MCP from protein sequences, utilizing the data set generated by physics-based computer simulations. However, our previous predictor relied on evolutionary information obtained from multiple sequence alignments (MSAs), which constrained the speed of the predictions. Here, we introduce a new transformer-based model, ProtRAP-LM, which leverages protein language model (LM) embeddings as input features to rapidly and accurately predict MCP for each residue within a given protein sequence. ProtRAP-LM demonstrates superior performance on a 184-protein test set of compared to the MSA-based model, achieving a speed-up of over 300 times on a workstation equipped with an RTX 3080 Nvidia GPU. As a result, entire proteomes can be predicted within hours, enabling us to provide more comprehensive annotations of membrane protein sequences on a proteome-wide scale, particularly for single-pass transmembrane proteins, membrane-anchored proteins, and β -sheet-containing membrane proteins, which have long-posed challenges in the field. In the end, we provide a comprehensive list of membrane proteins for 48 living organisms, offering a rich resource for investigating the structure and function of these essential biomolecules in the future. In particular, we propose 78 potential human membrane proteins that have not been previously identified. An online computation server for ProtRAP-LM is available at http://www.songlab.cn/ProtRAP-LM/home/.
Wang et al. (Sun,) studied this question.