s******y
发帖数: 28562 | 1 Sorry I cannot input Chinese today.
I have a question here. We had some old data for tumor growth on immune
compromised mice. When we did the experiments three years ago, we did two
different ways without thinking too much. In one way, we inject the test
cell and the control cell on the same mice (different side) in a paired
study. In another way, we inject the test cell and control cell on different
mice in a unpaired study.
When we analyze the data, we found that in the paired study, the test cells
always formed smaller tumor than the control cells injected on the same mice
. Therefore, the error bar is really good and the p value is excellent.
However, the n number for this group is smaller (n=5) because some mice died
in the course.
However, for the unpaired study, when the cells were injected in different
mice, the difference is not so clear because in a few out-liner mice where
the tumors for the test cells grow very, very big. Therefore, the error bar
is not very good. We suspect that when we injected the cells in different
mice, we unknowingly brought in the individual variation between different
mice. So now we are questioning whether we should show this group of data or
not? The n number for this group is 7.
So, for experts of mice xenograft, which dataset do you think we should use?
Also, do you think our n number is good enough? For the first method, even
though the n number is small, the p value is less than 0.01 because the
difference is very significant. But do you think the reviewer will have a
problem with the n number?
If we combine the two dataset together, we will get a n number of 12.
However the error bar doesn't look very good, but if we analyze the curve
with paired-t test, the p value is still OK. | s******y
发帖数: 28562 | 2 自己顶一下。有人知道么？肿瘤细胞需要和对照组一起打在同一个老鼠身上么？还是无
所谓乱打也可以？
error
【在 s******y 的大作中提到】
: Sorry I cannot input Chinese today.
: I have a question here. We had some old data for tumor growth on immune
: compromised mice. When we did the experiments three years ago, we did two
: different ways without thinking too much. In one way, we inject the test
: cell and the control cell on the same mice (different side) in a paired
: study. In another way, we inject the test cell and control cell on different
: mice in a unpaired study.
: When we analyze the data, we found that in the paired study, the test cells
: always formed smaller tumor than the control cells injected on the same mice
: . Therefore, the error bar is really good and the p value is excellent.
| r*****i
发帖数: 117 | 3 should be in the same mouse. Otherwise, how can you exclude the viability
among mice. | r*****i
发帖数: 117 | 4 check sean morrison's nature in 2012. regarding cancer stem cell. | s******y
发帖数: 28562 | 5 谢谢！
我们打在同一个老鼠身上的数据非常的好，对照组和实验组有很大的差别。但是乱打的
差别就会小一些。我再问一个傻问题，对于这种皮下肿瘤试验，n=? 是一个可以接受的
数值？ 我们有一个数据是n=6,另外一个数据是n=3 （汗，当时做这个试验的时候实在
是没有经验）。会不会被审稿者立刻把稿件扔我们脸上？
【在 r*****i 的大作中提到】
: should be in the same mouse. Otherwise, how can you exclude the viability
: among mice.
| s******y
发帖数: 28562 | 6 Thanks. Interesting paper.
【在 r*****i 的大作中提到】
: check sean morrison's nature in 2012. regarding cancer stem cell.
| r*****i
发帖数: 117 | 7 it depends on how familiar you and reviewers are, and how rare your samples
can be got.
Maybe you can explain this to the reviewers.
sometimes, we also have n=3.
【在 s******y 的大作中提到】
: 谢谢！
: 我们打在同一个老鼠身上的数据非常的好，对照组和实验组有很大的差别。但是乱打的
: 差别就会小一些。我再问一个傻问题，对于这种皮下肿瘤试验，n=? 是一个可以接受的
: 数值？ 我们有一个数据是n=6,另外一个数据是n=3 （汗，当时做这个试验的时候实在
: 是没有经验）。会不会被审稿者立刻把稿件扔我们脸上？
| s******y
发帖数: 28562 | 8 我们原先不是做癌症起家的，所以根本不认识癌症领域里的任何人。
样品其实并不rare，但是那个实验是好几年前我还在前老板那里的时候做的了。
现在时过境迁，要重复一下挺麻烦的，主要是需要时间去重新弄那些细胞和老鼠。n=6
的样品和n=3的样品都是做同一个蛋白的，唯一不同的是测的两个不同isoform, 但是结
论都是一样的。
samples
【在 r*****i 的大作中提到】
: it depends on how familiar you and reviewers are, and how rare your samples
: can be got.
: Maybe you can explain this to the reviewers.
: sometimes, we also have n=3.
| G********e
发帖数: 528 | 9 那得谨慎
下结论的时候保守一点可能问题不大
我觉得同一只老鼠也有caveat
万一实验组对照组在体内遥控打架怎么办
比较呆在体内好几个礼拜
不过你也有不是同一只老鼠的数据支持
两种证据加起来其实很有说服力
这些体内实验可以在文章里面downplay一点，如果你不想重复的话。
你不下能被人攻击的结论自然没有人攻击你
体外实验是重点，可以故意摆几个漏洞让审稿人挑
让他们挑的津津有味
回来你补上就行了
瞎扯一下，别当真
肿瘤最无聊了，不喜欢
6
【在 s******y 的大作中提到】
: 我们原先不是做癌症起家的，所以根本不认识癌症领域里的任何人。
: 样品其实并不rare，但是那个实验是好几年前我还在前老板那里的时候做的了。
: 现在时过境迁，要重复一下挺麻烦的，主要是需要时间去重新弄那些细胞和老鼠。n=6
: 的样品和n=3的样品都是做同一个蛋白的，唯一不同的是测的两个不同isoform, 但是结
: 论都是一样的。
:
: samples
| s******y
发帖数: 28562 | 10 谢谢。故意卖几个破绽这个建议很有新意啊，呵呵。
【在 G********e 的大作中提到】
: 那得谨慎
: 下结论的时候保守一点可能问题不大
: 我觉得同一只老鼠也有caveat
: 万一实验组对照组在体内遥控打架怎么办
: 比较呆在体内好几个礼拜
: 不过你也有不是同一只老鼠的数据支持
: 两种证据加起来其实很有说服力
: 这些体内实验可以在文章里面downplay一点，如果你不想重复的话。
: 你不下能被人攻击的结论自然没有人攻击你
: 体外实验是重点，可以故意摆几个漏洞让审稿人挑
|
|
