Simple consensus procedures are effective and sufficient in secondary structure prediction
Simple consensus procedures are effective and sufficient in secondary structure prediction
复制标题
DOI：
10.1093/protein/gzg063
复制
发表时间：
2003-07-01
期刊：
影响因子：
--
通讯作者：
Valle, G
中科院分区：
文献类型：
--
作者：
Albrecht, M;Tosatto, SCE;Valle, G
我们已经分析了多数表决的最小组合集的三个国家的最先进的二级结构预测方法，以获得一致的预测性能。使用来自伊娃服务器的三个大型基准集，我们的结果显示，通过形成共识，平均Q（3）预测准确率显著提高，最高可达1.5个百分点。对预测的二级结构元素应用额外的平凡过滤程序太短，不会显著影响预测准确度。我们的分析还提供了有价值的洞察力的预测方法，我们联合收割机的结果的相似性，以及一致预测的二级结构的置信度较高。
We have analyzed the performance of majority voting on minimal combination sets of three state-of-the-art secondary structure prediction methods in order to obtain a consensus prediction. Using three large benchmark sets from the EVA server, our results show a significant improvement in the average Q(3) prediction accuracy of up to 1.5 percentage points by consensus formation. The application of an additional trivial filtering procedure for predicted secondary structure elements that are too short, does not significantly affect the prediction accuracy. Our analysis also provides valuable insight into the similarity of the results of the prediction methods that we combine as well as the higher confidence in consistently predicted secondary structure.