Development of a prediction system for tail-anchored proteins

Area of Science:

• Cell Biology
• Proteomics
• Bioinformatics

Background:

• Tail-anchored (TA) proteins are a class of transmembrane proteins characterized by a C-terminal transmembrane domain (TMD) and lacking an N-terminal signal sequence.
• TA proteins play vital roles in cellular functions such as membrane fusion, apoptosis regulation, and vesicular transport, constituting 3-5% of all transmembrane proteins.
• Predicting TA proteins typically requires a combination of TMD and signal sequence prediction tools.

Purpose of the Study:

• To develop a novel prediction system for identifying tail-anchored (TA) proteins.
• To enable TA protein prediction solely from amino acid sequences, simplifying the identification process.

Main Methods:

• Developed the Tail-Anchored Protein Prediction Model (TAPPM) utilizing sequence features of TMDs and flanking regions of TA proteins.
• Collected manually curated TA proteins from published literature.
• Constructed Hidden Markov Models (HMMs) for TA proteins and three other transmembrane protein types with similar structures to compare their likelihoods.

Main Results:

• The TAPPM system demonstrated high prediction accuracy for TA proteins.
• Achieved an area under the receiver operator curve (AUC) value of 0.963, indicating robust performance.
• Developed a command-line tool in Python for practical application of the prediction system.

Conclusions:

• The developed HMM models and TAPPM system provide a highly accurate method for predicting TA proteins.
• The availability of a command-line tool facilitates the broader application of this prediction method in research.
• This advancement simplifies the identification of TA proteins, supporting further research into their functions.