A single-residue affinity scale for DNA-binding using linear perceptron

A linear scale to estimate DNA-binding free energy of amino acid residues is reported. Scales derived exclusively for irregular and helical positions give 76% and 68% classification accuracy between stabilizing and destabilizing protein-DNA interaction. Mean absolute error (MAE) in ddG values is 0.786 and 0.883 kcal/mol respectively. Without using structure information of residues to derive affinity scales, 67.0% mutations could be correctly classified between those stabilizing and destabilizing binding. Mean absolute error (MAE) and correlation of ddG predictions are 0.953 kcal/mol and 0.385 respectively. PRIB 2008 proceedings found at: http://dx.doi.org/10.1007/978-3-540-88436-1 Contributors: Monash University. Faculty of Information Technology. Gippsland School of Information Technology ; Chetty, Madhu ; Ahmad, Shandar ; Ngom, Alioune ; Teng, Shyh Wei ; Third IAPR International Conference on Pattern Recognition in Bioinformatics (PRIB) (3rd : 2008 : Melbourne, Australia) ; Coverage: Rights: Copyright by Third IAPR International Conference on Pattern Recognition in Bioinformatics. All rights reserved.本研究报道了一种可用于估算氨基酸残基DNA结合自由能的线性标度。仅针对不规则构象位点与螺旋构象位点所衍生的两类标度，在区分稳定与去稳定蛋白质-DNA相互作用时，分类准确率分别为76%与68%；两类标度对应的ddG值的平均绝对误差（mean absolute error, MAE）分别为0.786与0.883 kcal/mol。若不借助残基结构信息来推导亲和力标度，仍可将67.0%的突变正确划分为稳定结合型与去稳定结合型；此时ddG预测值的平均绝对误差与相关系数分别为0.953 kcal/mol与0.385。相关研究内容收录于2008年第三届IAPR国际模式识别生物信息学会议（PRIB 2008）论文集，获取地址：http://dx.doi.org/10.1007/978-3-540-88436-1。贡献项包括：莫纳什大学（Monash University）信息技术学院吉普斯兰信息技术分校；作者团队包括Chetty, Madhu、Ahmad, Shandar、Ngom, Alioune、Teng, Shyh Wei；该数据集源自第三届IAPR国际模式识别生物信息学会议（PRIB 2008，2008年于澳大利亚墨尔本举办）。数据集版权归第三届IAPR国际模式识别生物信息学会议所有，保留所有权利。