SaProtHub/Dataset-DLG4_RAT

该数据集包含蛋白质DLG4_RAT氨基酸序列的单点突变及其在深度突变扫描实验中对应的突变效应分数。蛋白质序列以单一氨基酸序列的格式表示。数据集分为训练集、验证集和测试集，分别包含1325、159和176个样本。标签代表每个突变氨基酸序列的适应度分数，基于观察到的每个氨基酸在选定的种群与未选定的种群中的频率的对数，相对于野生型。效应分数范围从负无穷大到正无穷大，分数大于0表示高适应度，小于0表示低适应度。

This dataset contains single site mutations of the protein DLG4_RAT amino acid sequence and their corresponding mutation effect scores from a deep mutation scanning experiment. The protein sequences are represented in the format of a single amino acid sequence. The dataset is split into training, validation, and test sets, containing 1325, 159, and 176 samples respectively. The labels represent the fitness score of each mutant amino acid sequence based on the frequency of observing each amino acid at each position in the selected versus the unselected population, relative to the wild type. The effect scores range from negative infinity to positive infinity, with scores greater than 0 indicating high fitness and scores less than 0 indicating low fitness.