Overton Ian M, Barton Geoffrey J
School of Life Sciences, University of Dundee, Dow Street, Dundee DD1 5EH, UK.
FEBS Lett. 2006 Jul 10;580(16):4005-9. doi: 10.1016/j.febslet.2006.06.015. Epub 2006 Jun 16.
Target selection and ranking is fundamental to structural genomics. We present a Z-score scale, the "OB-Score", to rank potential targets by their predicted propensity to produce diffraction-quality crystals. The OB-Score is derived from a matrix of predicted isoelectric point and hydrophobicity values for nonredundant PDB entries solved to <or=3.0 A against a background of UniRef50. A highly significant difference was found between the OB-Scores for TargetDB test datasets. A wide range of OB-Scores was observed across 241 proteomes and within 7868 PfamA families; 73.4% of PfamA families contain >or=1 member with a high OB-Score, presenting favourable candidates for structural studies.
靶点选择和排序是结构基因组学的基础。我们提出了一种Z分数标度,即“OB分数”,用于根据预测的产生衍射质量晶体的倾向对潜在靶点进行排序。OB分数源自一个矩阵,该矩阵基于UniRef50背景,针对解析度≤3.0 Å的非冗余PDB条目预测其等电点和疏水性值。在TargetDB测试数据集的OB分数之间发现了高度显著的差异。在241个蛋白质组和7868个PfamA家族中观察到了广泛的OB分数范围;73.4%的PfamA家族包含≥1个具有高OB分数的成员,为结构研究提供了有利的候选对象。