生物信息学
生物信息學
생물신식학
China Journal of Bioinformatics
2015年
3期
165-169
,共5页
高光芹%黄家荣%周俊朝%谢鹏芳
高光芹%黃傢榮%週俊朝%謝鵬芳
고광근%황가영%주준조%사붕방
小黑杨%磷酸化蛋白质%磷酸化位点%人工神经网络
小黑楊%燐痠化蛋白質%燐痠化位點%人工神經網絡
소흑양%린산화단백질%린산화위점%인공신경망락
Populus simonii×P nigra%Phosphoproteome%Phosphorylation site%Artificial neural network
以小黑杨磷酸化蛋白质组为研究对象,用人工神经网络表达丝氨酸、苏氨酸等残基位点的磷酸化与氨基酸序列的结构特征之间的非线性关系,建立了BP 人工神经网络模型,并用磷酸化数据对所建模型进行训练和分析,得适宜的结构为21×16 : 8 : 4,拟合准确度为90%,Acc、Sn、Sp、MCC分别为78%、89%、67%、0.57,对比分析结果表明,所建模型具有较强的预测能力.
以小黑楊燐痠化蛋白質組為研究對象,用人工神經網絡錶達絲氨痠、囌氨痠等殘基位點的燐痠化與氨基痠序列的結構特徵之間的非線性關繫,建立瞭BP 人工神經網絡模型,併用燐痠化數據對所建模型進行訓練和分析,得適宜的結構為21×16 : 8 : 4,擬閤準確度為90%,Acc、Sn、Sp、MCC分彆為78%、89%、67%、0.57,對比分析結果錶明,所建模型具有較彊的預測能力.
이소흑양린산화단백질조위연구대상,용인공신경망락표체사안산、소안산등잔기위점적린산화여안기산서렬적결구특정지간적비선성관계,건립료BP 인공신경망락모형,병용린산화수거대소건모형진행훈련화분석,득괄의적결구위21×16 : 8 : 4,의합준학도위90%,Acc、Sn、Sp、MCC분별위78%、89%、67%、0.57,대비분석결과표명,소건모형구유교강적예측능력.
In this paper, the phosphoproteome of Populus simonii × P nigra was used as the research object. The nonlinear relationship between the structure characteristics of amino acid sequence and phosphorylation of serine and threonine was expressed by artificial neural network. A BP artificial neural network model was established and trained by using the real data on phosphorylation. The appropriate structure is 21 x 16 : 8 : 4, the fitting accuracy is 90%, and the Acc, Sn, Sp, MCC are 78%, 89%, 67%, and 0. 57, respectively. The comparative results show that the model has strong prediction ability.
