中洪博元,肿瘤动物,实验分组 z-bo 2020-07-21 1079

The experimental grouping of experimental animals is a very important part of the experimental design. If required by the American Animal Committee, it can provide basic animal information, such as animal source, strain, living environment, grouping method, etc. However, most of the published articles did not describe the method of random grouping of experimental animals, and the editors required original data to prove the reliability of the data. You will see that grouping is very important. So what principles should be emphasized when grouping laboratory animals, and how to achieve them? The grouping of animals should generally follow the principles of control and randomization. There are two main problems. The first is how to design the control group. So how can animals be randomly assigned to these groups?

　　 Question 1: How to establish a control group?

　　 To ensure that you get scientific and effective data, the first design principle of grouping is to set up a control group. Generally, to observe the influence of drugs or certain external factors on experimental animals, please set up three groups of control groups: blank control group (negative control group), model control group (positive control group) and positive control group. Blank control group: refers to experimental animals that have not received any treatment, the purpose is to emphasize the difference between normal animals and diseased animals. If the therapeutic effects of certain drugs on diabetic mice are being studied, the group is considered unaffected. Diabetic wild-type mice can receive normal saline or other drug media as needed. I can do it. Model control group: an animal model of disease that has undergone negative treatment. The purpose is to observe the effect of the drug and the treatment group under medical conditions and animal behavior. If you want to see whether a specific drug has a therapeutic effect on breast cancer, the animals in the treatment group need to be administered intraperitoneally, and the model group should use the same dose of drug-free solvent, such as saline. Will be injected. Positive control group: usually given disease model animals known to be effective drugs or other effective factors; the purpose is to compare whether the test drug is effective with the positive drug. Example: Doxorubicin is known to have an inhibitory effect on the growth of breast cancer tumors, so doxorubicin Can be used as a positive control to determine the efficacy of the study drug.

　　 Next, how many groups in the control group are usually required for your experimental design? In the experimental design, in addition to the above-mentioned setting of the control group, a treatment factor group is also required, but it is usually best to set a dose gradient and design 3 or more to produce a dose effect. In summary, the recommended groups are: ① blank group, ② model group, ③ model + positive drug model, ④ model + low dose drug, ⑤ model + medium dose drug, ⑥ model + high dose drug.

　　 Note: If the subject does not need medical conditions, the model group is not needed;

　　 Note: Not all experiments can find positive control drugs, so all experiments need positive control

　　 Note: For specific dosage design, please refer to the link at the end of the article (drug concentration design)

　　 Question 2: What is the correct number for each group?

　　The design of animal experiments must follow the implementation of the "3R principle". This includes the principles of substitution, reduction and optimization of laboratory animals. Reduction means reducing the number of laboratory animals. Checking the literature did not find that there is an absolute requirement for the number of experimental animals, but the number must be reduced while still meeting the statistical requirements. To understand statistics, each group requires at least 6 available data.

　　 There are usually more than 10 ordinary mice per group.

　　 There are 6 or more ordinary rats in each group.

　　 The higher the number of large animals, the higher the price, and the price will be appropriately lowered according to the situation, but it is usually not a minority. Only 4-5,

　　 Question 3: Please tell me how to allocate randomly.

　　Random allocation can ensure the scientificity of the data and ensure that the data is more attractive. Therefore, the experimental animals should be grouped strictly according to the randomization principle, so that each animal has the same chance of being assigned to each experimental group. Fundamentally different experimental goals, experimental goals and commonly used grouping methods are completely random design, random block design, etc. In order to realize random number grouping, random number table and random number residual grouping method are needed.

　　What is a random number table? Also known as a random number table, it is a table of 10 numbers from 0 to 9 randomly generated by a computer. The probability of each number appearing in the table is basically the same. When grouping, you need to borrow this tool to achieve random allocation.

　　Method 1: Completely random design

　　 Randomly group animals into random numbers generated by a table or computer;

　　Scope: All animals are "homologous" or close to "All animals must be "homologous", that is, the sex, weight and other related factors (such as tumor burden) of all animals are the same or similar.

　　Description:

　　Numbering: Experimental animals are usually numbered from 1 to N according to their body weight.

　　Get random numbers: Start from any number in the random number table, and get N random numbers in the same direction. digital;

　　 Find the remainder: Divide the random number by the group number to find the remainder. If split, the rest is the number of groups.

　　Group: Group by the remaining part.

　　Adjustment: If there is no equal distribution, please continue to copy the random numbers in order. The divisor will be the number a of the largest group number, and the resulting remainder will be used as the serial number to be drawn until it is adjusted to be equal (if it is divided equally, the remainder will be a). ..

　　 For example, 15 mice are divided into 3 groups. First, number the animals according to their body weight, then use the random table to replicate 15 numbers in succession, and then use group number 3 to find the remaining animals. In other words, dividing by 3 and the rest being 1, 2, and 3 means entering groups A, B, and C respectively. However, there is only one in Group A (6). Check the random number table. The next random number is 55.55/6 greater than 1, so the first case is adjusted from group A to group C.

　　Advantages: simple operation, can realize the principle of random allocation;

　　 Disadvantages: First of all, the process is cumbersome, usually one operation cannot make the number of groups equal, and needs to be adjusted; too many groups are not easy to quickly achieve grouping. This method is not suitable for larger weight ranges.

　　Method 2: Random block design

　　 However, if you have a small group of 18g and a large pile of 30g, the weight difference will be relatively large. When assigned to each group according to the completely random method described above, the weight components are particularly large, and some components are particularly small. Because it is difficult to guarantee uniform distribution, random block design was born.

　　 Random block design: called balanced randomization or finite randomization. In other words, first divide a certain number and percentage of animals into several blocks (usually → 3). It is expected that the length of pieces such as animals is greater than twice the number of groups, and if they are too small, they tend not to be random. If you have 4 groups, it is recommended that you choose 8 or more.

　　 Example: There are 40 SD rats weighing between 180 and 220 grams. It should be divided into 4 groups, namely 10 rats in each group (A, B, C, D). The animals were divided into several groups according to their body weight, and then the animals in each group were randomly assigned to each group.

　　 First, weigh the mice and sort them.

Partition group: try to divide these rats into 4 blocks according to their body weight (assuming 180-190g/8 rats, 190-200g/12 rats, 200-210g/16 rats, 210-220g/4 Rats), that is, 4 liters, the number of each bathroom is at least more than 4 when separated;

　　#: 4 liters are numbered sequentially, the first bathroom is 1, 2, 3, 4...8, and so on. The number is up to 40.

　　Random allocation block group 1: Randomly allocate to these 8-4 groups. There are two in each group. The grouping method uses the above random number table and allocates according to the remaining random numbers. The remaining 3 blocks

　　3 are randomly assigned to each group in the same way as above.

　　Advantages: interval grouping can minimize the biological differences between experimental groups, comply with the principles of consistency and randomness, and make the experimental results more statistically meaningful.

　　 Note: Sometimes multiple factors will be encountered and need to be decomposed in order. For example, there are currently 40 male and female mice with large differences in body weight. Not only should they be allocated according to their weight, but the males and females should also be separated first. Separate and layer by weight. In other words, it must be randomly divided into equal groups based on a certain element, and divided into certain areas within the group based on another element.

　　 Note: If the number is small, you can manually calculate the grouping, but if the number is too large, you need to use software such as Excel or SPSS to avoid the trouble of checking the random number table and random numbers. The job will be processed in a few minutes. All in all, we found that the grouping of experimental animals requires compliance with the principles of control and randomness to ensure that the data is more scientific and statistically significant. It is best not to add subjective elements to the group. As an important part of experimental design, the method of grouping experimental animals should be carefully considered before starting the experiment. Through our article, we hope that everyone understands the importance of random grouping and can actually learn random grouping.