Detection of circular permutations by Protein Language Models

23 Apr 2024  ·  Yue Hu, Bin Huang ·

Circular permutation in proteins is a phenomenon with significant implications in the study of protein evolution and functionality. The detection of such permutations is traditionally reliant on structural alignment methods, which, while accurate, are computationally intensive and require known three-dimensional structures. The circular permutation detection method based on sequence dependencies may be faster, but it may sacrifice some accuracy, especially for structures with low sequence similarity. To address these limitations, we introduce a novel protein language model, the plmCP, designed to detect circular permutations directly from amino acid sequences. The protein language model can help us extract structural information from sequences. Utilizing advanced embedding techniques, our model infers structural information from the linear sequence data, enabling the identification of permuted proteins with a high degree of accuracy. In case study, this model has a good result compared with traditional structural alignment methods in speed and accessibility, requiring only amino acid sequences to deliver results comparable to established techniques. In conclusion, our protein language model presents a promising tool for the efficient and accessible detection of circular permutations in proteins, with the potential to facilitate advancements in the understanding of protein evolution and engineering. The source code can be obtained from https://github.com/YueHuLab/plmCP.

PDF Abstract

Datasets


  Add Datasets introduced or used in this paper

Results from the Paper


  Submit results from this paper to get state-of-the-art GitHub badges and help the community compare results to other papers.

Methods