R语言S4对象

创建类

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
setClass(
  # 类名
  "TumorPrognosis", 
  
  # 数据列表
  slots = c(
    rawdata = "data.frame",
    data = "data.frame",
    markers = "character",
    survival_year = "numeric",
    group = "character",
    cutoff = "list",
    roc = "list",
    ROCPlot = "list",
    KMCurve = "list"
  )
)

构造函数
initialize是内置的构造函数,
这也是范式方法的增加方法，如果是范式方法，可以直接给S4类增加这个方法

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
setMethod("initialize", signature(.Object = "TumorPrognosis"),
          function(.Object, tumorName, markers, survival_year){
            .Object@markers = markers
            .Object@survival_year = survival_year
            .Object@rawdata = read.xlsx("01.data/prognosis1.xlsx", sheet = tumorName)
            
            colnames(.Object@rawdata) = c(
              "ID",
              "inTime",
              "outInfo",
              "stage",
              markers
            )
            
            .Object = ScreenData(.Object)
            .Object = GroupedSample(.Object)
            .Object = CalcBestCutoff(.Object)
            .Object = getKMCurve(.Object)
            
            .Object
})

如果不是范式，就需要先创建一个范式的，然后再添加到S4类上

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
setGeneric("getKMCurve", function(.Object) standardGeneric("getKMCurve"))
setMethod("getKMCurve", signature(.Object = "TumorPrognosis"), function(.Object){
  
  for (data_set_name in c("train", "valid")){
    
    
    
    
    sur_data = getGroupSameples(.Object, data_set_name)
    for (marker in c("stage", .Object@markers)){
      if(marker == "stage"){
        legend_labs = c("Stage I/II", "Stage III/IV")
      }else{
        legend_labs = paste0(marker, c(" high risk", " low rish"))
      }
      
      group = paste0(marker, "_b")
      surv = Surv(sur_data$futime, sur_data$fustat2)
      formula_this = as.formula(paste0("surv", "~", group))
      
      ggl = surv_fit(formula_this, data = sur_data)
      surv_diff = survdiff(formula_this, data = sur_data)
      p = ggsurvplot(ggl,
                     # palette = c("blue", "red"),
                     xlab = "Time (Years)",
                     ylab = "OS (percentage)",
                     break.x.by = 365,
                     break.y.by = 0.25,
                     ylim = c(0, 1),
                     xlim = c(0, 365 * .Object@survival_year),
                     legend.title = paste0("Log-rank p = ", format(1 - pchisq(surv_diff$chisq, length(surv_diff$n) -1), digits = 3)),   # 图例标题
                     legend.labs = legend_labs,
                     size = 0.8,
                     # pval = T,
                     # pval.size = 5,
                     # pval.coord=c(365,0.05),
                     censor.shape=124,
                     censor.size=2,
                     font.main = c(16, "plain", "darkblue"),
                     font.x = c(10, "plain", "black"),
                     font.y = c(10, "plain", "black"),
                     font.tickslab = c(8,"plain", "black"),
                     font.legend = c(8,"plain","black"),
                     risk.table =F,
                     tables.height = 0.25,
                     # legend.labs=c("LPA","Non-LPA"),
                     legend = c(0.3,0.28),
                     fontsize=4.5,
                     conf.int=T,
                     conf.int.style='ribbon',
                     conf.int.alpha=0.1
                     # surv.median.line="hv"
      )
      
      p = p$plot +
        scale_x_continuous(breaks = seq(0, 1825, 365), labels = seq(0, 5))
      
      .Object@KMCurve[[data_set_name]][[marker]] = p
    }
  }
  .Object
  
})

创建类实例

1
T05_5 = new("TumorPrognosis", "T05", c("CEA", "CA199", "AFP")  , 5)