R实战

masigpro

跟着Cell学作图 | 6.时间序列分析(Mfuzz包)

一个答疑教程。

maSigPro 流程 示例数据 #BiocManager::install(‘maSigPro’)library(maSigPro)# 载入示例数据data(data.abiotic) data.abiotic[1:5,1:5]data(edesign.abiotic)head(edesign.abiotic) > data.abiotic[1:5,1:5]        Control_3H_1 Control_3H_2 Control_3H_3 Control_9H_1 Control_9H_2STMDF90   0.13735714   -0.3653065  -0.15329448   0.44754535  0.287476796STMCJ24           NA           NA           NA           NA           NASTMJH42   0.07864449    0.1002328  -0.17365488  -0.25279484  0.184855409STMDE66   0.22976991    0.4740975   0.46930716   0.37101059 -0.004992029STMIX74   0.14407618   -0.4801864  -0.07847999   0.05692331  0.013045420 > head(edesign.abiotic)             Time Replicate Control Cold Heat SaltControl_3H_1    3         1       1    0    0    0Control_3H_2    3         1       1    0    0    0Control_3H_3    3         1       1    0    0    0Control_9H_1    9         2       1    0    0    0Control_9H_2    9         2       1    0    0    0Control_9H_3    9         2       1    0    0    0

注意：data.abiotic是已经标准化过的基因表达矩阵。

建立回归模型 生成回归矩阵(makeDesignMatrix) design  design$groups.vector [1] “ColdvsControl” “HeatvsControl” “SaltvsControl” “Control”       [5] “ColdvsControl” “HeatvsControl” “SaltvsControl” “Control”       [9] “ColdvsControl” “HeatvsControl” “SaltvsControl” 寻找重要基因(p.vector)

F检验确定回归方程的显著性，采用BH的校正方式，校正多重假设检验的p值。

校正后的p值小于p.vector的参数Q的基因就作为候选基因，进行下一步的分析。通过fit$SELEC可以得到候选基因的表达量信息。

fit