Micsonai András, Wien Frank, Murvai Nikoletta, Nyiri Márton Péter, Balatoni Bori, Lee Young-Ho, Molnár Tamás, Goto Yuji, Jamme Frédéric, Kardos József
Department of Biochemistry, Institute of Biology, ELTE Eötvös Loránd University, Budapest H-1117, Hungary.
ELTE-Functional Nucleic Acid Motifs Research Group, Department of Biochemistry, Institute of Biology, ELTE Eötvös Loránd University, Budapest H-1117, Hungary.
Nucleic Acids Res. 2025 Jul 7;53(W1):W73-W83. doi: 10.1093/nar/gkaf378.
Circular dichroism (CD) spectroscopy is a widely used technique to characterize the secondary structure composition of proteins. We have developed the Beta Structure Selection (BeStSel) method (PNAS, 112, E3095), which solves the main problem of protein CD spectroscopy-namely, the spectral variability of β-structures. The BeStSel web server utilizes this method to provide tools to the community for CD spectrum analysis. BeStSel uniquely provides information on eight secondary structure components, including parallel β-structure and antiparallel β-sheets with three different twist groups. It outperforms all available methods in accuracy and information content, and is also able to predict protein folds down to the topology/homology level of the CATH classification. The algorithm has been further developed, and the accuracy of the estimation of the secondary structure elements is improved by 0.7% as an average on the reference dataset. A new module of the web server calculates protein stability from the thermal denaturation profile followed by CD. Secondary structure calculations of uploaded PDB and mmcif files support the experimental verification of MD simulations and AlphaFold models by CD spectroscopy. Well-proven modules for disorder-order classification and extinction coefficient calculation continue to work. The BeStSel server is freely accessible at https://bestsel.elte.hu.
圆二色性(CD)光谱是一种广泛用于表征蛋白质二级结构组成的技术。我们开发了β结构选择(BeStSel)方法(《美国国家科学院院刊》,112,E3095),该方法解决了蛋白质CD光谱的主要问题,即β结构的光谱变异性。BeStSel网络服务器利用此方法为社区提供用于CD光谱分析的工具。BeStSel独特地提供了有关八种二级结构成分的信息,包括具有三种不同扭曲组的平行β结构和反平行β折叠。它在准确性和信息含量方面优于所有现有方法,并且还能够预测低至CATH分类的拓扑/同源性水平的蛋白质折叠。该算法已得到进一步开发,在参考数据集上,二级结构元素估计的准确性平均提高了0.7%。网络服务器的一个新模块根据随后进行CD的热变性曲线计算蛋白质稳定性。上传的PDB和mmcif文件的二级结构计算支持通过CD光谱对分子动力学模拟和AlphaFold模型进行实验验证。经过充分验证的无序-有序分类和消光系数计算模块继续发挥作用。可通过https://bestsel.elte.hu免费访问BeStSel服务器。