半数致死量可以这样计算么？

是 否 +2 论坛币

k人 参与回答

经管之家送您一份

应届毕业生专属福利!

求职就业群

赵安豆老师微信：zhaoandou666

经管之家联合CDA

送您一个全额奖学金名额~ !

立即领取

感谢您参与论坛问题回答

经管之家送您两个论坛币！

+2 论坛币

我现在想通过R计算半数致死量，数据如下：
conc           nlive
1000000      0
100000        1
10000          8

conc列是毒物浓度，nlive为存活的动物数。每种剂量的动物总数为10只，总共三个浓度。

conc<-c(1000000,100000,10000)
dose<-log10(conc)
numlive<-c(0,1,8)
total<-cbind(numlive,10-numlive)
fit<-glm(total~dose,family = binomial)
library(MASS)
tmp<-dose.p(fit,p=0.5)
logld50<-tmp[1]
SE<-attr(tmp,"SE")
ld50<-10**logld50
ci<-10**(logld50+c(-1.96,1.96)*SE)
logld50
ld50
ci

参考了文献后设计了以上代码计算ld50，及其95%置信区间。但我想画一个以剂量为横坐标，以死亡率为纵坐标的“S”型曲线。这该如何实现呢？

扫码加我 拉你入群

请注明：姓名-公司-职位

以便审核进群资格，未注明则拒绝

分享0 收藏0 回帖

关键词：live con 动物

或者还有其他R里的函数用来计算半数致死量么？

不明白什么是半数致死量

1. # For plotting
   
2. newdose <- seq( from = min(dose), to = max(dose), length.out = 101 )
   
3. newprob <- predict( fit, newdata = data.frame( dose = newdose ), type = "response" )
   
4. plot( newdose, newprob, type = "l", xlab = "log10(conc)", ylab = "p(death)" )

复制代码

@suzhzh:
According to wikipedia:

The median lethal dose, LD50 (abbreviation for "lethal dose, 50%"), LC50 (lethal concentration, 50%) or LCt50 (lethal concentration and time) of a toxin, radiation, or pathogen is the dose required to kill half the members of a tested population after a specified test duration. LD50 figures are frequently used as a general indicator of a substance's acute toxicity. A lower LD50 is indicative of increased toxicity.

zhangyangsmith 发表于 2017-2-13 17:39
@suzhzh:
According to wikipedia:

谢谢。代码没弄明白，您能给解释一下么？另外，能否画出半数致死量所在的水平线和垂线呢？

lanhong1993 发表于 2017-2-14 13:44
谢谢。代码没弄明白，您能给解释一下么？另外，能否画出半数致死量所在的水平线和垂线呢？

1. # Following your original scripts
   
2. 
   
3. # Create a serial of independent variables (101 data points) within the range of source data
   
4. newdose <- seq( from = min(dose), to = max(dose), length.out = 101 )
   
5. 
   
6. # Compute the probability (dependent variables) corresponding to the independent variables created above
   
7. newprob <- predict( fit, newdata = data.frame( dose = newdose ), type = "response" )
   
8. 
   
9. # Plot probability (dependent variables) against log-conc (independent variable)
   
10. plot( newdose, newprob, type = "l", xlab = "log10(conc)", ylab = "p(death)" )
    
11. 
    
12. # Add referencing line
    
13. 
    
14. ## Horizontal
    
15. lines( x = c( par("usr")[1], logld50 ), y = c(0.5, 0.5), lty = 4 )
    
16. 
    
17. ## Vertical
    
18. lines( x = c(logld50, logld50), y = ( par("usr")[3], 0.5 ), lty = 4 )

复制代码