MOESM1 of Prediction of sgRNA on-target activity in bacteria by deep learning

Additional file 1 More detailed tables. Statistical significance of Spearman correlation coefficients in this paper and more details about 5-fold cross-validation. Table S1. Performance of all modes including DeepCRISPR, CNN_Lin, DeepCas9, CNN_2layers, CNN_3layers, CNN_4layers, CNN_5layers, and CNN_7layers. Table S2. The Spearman correlation coefficients and significances between eukaryotic sgRNA activity and predictions based on our prokaryotic models and eukaryotic models. Table S3. Significance between on-target activity and six melting temperatures, four RNA fold scores, and four POSs. Table S4. Detailed information of 5-fold cross-validation for several network architectures.

补充材料1：更详细的表格。包含本文中斯皮尔曼（Spearman）相关系数的统计学显著性，以及关于5折交叉验证的更多细节。表S1：所有模型的性能表现，涵盖DeepCRISPR、CNN_Lin、DeepCas9、CNN_2layers、CNN_3layers、CNN_4layers、CNN_5layers及CNN_7layers。表S2：真核生物单向导RNA（sgRNA）活性与基于本研究原核模型及真核模型所得预测结果之间的斯皮尔曼相关系数及其显著性。表S3：靶标活性与6种解链温度、4种RNA折叠得分及4种POSs之间的显著性关联。表S4：多种网络架构的5折交叉验证详细信息。