Help page
Randmice is an online tool that reduces group heterogeneity for in vivo experiments.
It allows you to distribute animals into groups in an optimal way, minimizing heterogeneity between groups. It can work with one or multiple covariates to optimize.
Also, it provides:
It allows you to distribute animals into groups in an optimal way, minimizing heterogeneity between groups. It can work with one or multiple covariates to optimize.
Also, it provides:
- The possibility to use less animals per experiment with the same heterogeneity;
- A tool to standardize randomization between experiments;
- Comply with ethical regulations (European directive n°2010/63/EU);
- Have an exhaustive report for our archive;
- One place to archive all your randomizations to keep an history.
Manual or random group balancing of animals is not optimal.
- Pure random assignment can lead to important heterogeneity between groups, especially true for small experiments (N < 10) [See Bertsimas et al. 2015].
- Manual balancing by operator's unique algorithm is not optimal either. First, operator cannot test all combinations (>10^10 for a 10 animals per group study with 10 groups). Also, it is based on its unique algorithm and introduces its unique biais, and so is not reproducible from one operator to another.
- Increasing the number of animal per group to reduce heterogeneity in pure random assigment does not help to comply with ethical regulations (3Rs = Reducing and European Regulations n°2010/63/UE / n°2019/1010).
Randmice uses bruteforce algorithms to distribute animals, calculates the heterogeneity, and iterates until an optimal distribution is found.
- Enter your data: Fill in or import your experimental data (animal ID, variables, etc.), choose the number of groups and the algorithm power.
- Balancing: Run the optimization with randmice. The calculation takes from 1 to 10 minutes depending on the number of animals and iterations.
- Results: Results are sent in a report by email. The report is also available in your dashboard if you have an account.
Randmice allows you to :
-
Reduce heterogeneity between groups compared to randomization.
For example, in an experiment where each mouse has two tumors of different sizes, randmice reduces heterogeneity by 97% compared to manual randomization. - Standardizethe group assignment proccess so you have reproductibility from one experiment to another
Yes, randmice allows you to reduce the number of animals per group while maintaining low heterogeneity.
For example, moving from 8 to 6 mice per group saves 2 mice per group without loss of statistical quality by using randmice.
For example, moving from 8 to 6 mice per group saves 2 mice per group without loss of statistical quality by using randmice.
Yes, randmice helps comply with the 3Rs rule (Replacement, Reduction, Refinement)
This ethical guideline is required by European directive n°2010/63/EU . It is adapted to use the smallest possible number of animals in a experiment, and so fully comply with rule reduction.
This ethical guideline is required by European directive n°2010/63/EU . It is adapted to use the smallest possible number of animals in a experiment, and so fully comply with rule reduction.
Yes, for example:
Jneid et al., Selective STING stimulation in dendritic cells primes antitumor T cell responses. Science Immunology.
This is important for us to be cited in material & methods, so please keep us informed when a paper is published using randmice.
See instructions on how to cite randmice.
Jneid et al., Selective STING stimulation in dendritic cells primes antitumor T cell responses. Science Immunology.
This is important for us to be cited in material & methods, so please keep us informed when a paper is published using randmice.
See instructions on how to cite randmice.
The best place to cite randmice is in the material & methods section of your paper.
Please see an example from Jenid et al. (2023):
Please see an example from Jenid et al. (2023):
Tumor-bearing mice were randomized using randmice (https://randmice.com) on the basis of tumor volume to distribute mice and homogenize the average tumor volume within the different groups.The algorithm randomly shuffles all mice between the groups and calculates the average tumor volume for each group. A total of 10^9 iterations were performed to minimize the difference in tumor volume average between all groups.
Yes, you can try randmice for free.
You can also create an account for free to keep track of your balancing runs and access them later.
Feel free to give us your feedback below!
You can also create an account for free to keep track of your balancing runs and access them later.
Feel free to give us your feedback below!
No, to use randmice you do not need an account.
The result of your balancing run will be sent by email when it is ready.
However, creating an account will help to keep track of all the balancing runs that will be performed. It will also help us to track activity and usage of our tool, to keep improving. If you want, create an account.
The result of your balancing run will be sent by email when it is ready.
However, creating an account will help to keep track of all the balancing runs that will be performed. It will also help us to track activity and usage of our tool, to keep improving. If you want, create an account.
Yes, fully free.
The idea of optimizing animal group-balancing came from our work at the laboratory and we believe it can benefit to other people. Our first version was very CPU expensive but we optimized our code to divide our cost by x1000 which now make it eligible for free usage. For now, the idea is to see how people are using it.
The idea of optimizing animal group-balancing came from our work at the laboratory and we believe it can benefit to other people. Our first version was very CPU expensive but we optimized our code to divide our cost by x1000 which now make it eligible for free usage. For now, the idea is to see how people are using it.
Any question?