Abstract
Nanobodies, also known as single domain antibodies or VHHs, are the artificial recombinant variable domains of heavy-chain-only antibodies. Nanobodies possess many unique properties, such as small size, good solubility, superior stability, rapid clearance from blood and deep tissue penetration. Therefore, nanobodies have emerged as promising tools for the diagnosis and treatment of diseases.
In recent years, many deep-learning based protein structure prediction methods have emerged that require only protein sequences input to obtain predictions of 3D protein structures with high credibility. Among them, AlphaFold2, RoseTTAFold, DeepAb, NanoNet and tFold performed excellently in the field of protein prediction or antibody/nanobody prediction. In this paper, we selected popular algorithms such as AlphaFold2, RoseTTAFold, DeepAb, tFold and NanoNet, and compared their performance on the prediction of 3D folded structure conformation of nanobody proteins.
In order to compare the performance of several algorithms in the prediction of nanobody protein structure, we selected 60 samples with known experimental 3D structures in the PDB database, and used these five prediction methods to predict their 3D structure from 2D amino acid sequences. After the prediction data is obtained, these dry lab prediction results are compared with the wet lab experiment results in the PDB database one by one, and finally the results are analyzed and discussed.
Competing Interest Statement
The authors have declared no competing interest.