Sometimes, it is difficult to validate mathematical model and computer simulation with real biological data? For example, it is impossible to do the correspond biological test for some wet-lab data. Is there another way that we can validate our model rather than comparison to real biological data?
In my opinion, if it's impossible to validate model results with real biological data or propose testable model predictions this model is no longer a science but something supernatural. By very definition science is the knowledge about the universe in the form of testable explanations and predictions. I think you may want to consider revise/simplify the model to link it to so some sort of experimental data or propose testable predictions for future experimental work.
Best, Anatoly
I agree with Anatoly. Every mathematical model is constructed based on biological, physical or chemical principles. These must be verified experimentally!
Even more, if your model depends on parameters then you must use biological data to determine their values.
Sometimes you can make some qualitative conclusions. Still I think that´s pretty far to be a complete research.
Sometimes the simulation is used to generate control testing and assume that the rest of the tests or simulations are compared with your data control.
In order to validate a biological mathematical model you really need real biological data, otherwise "validation" would have no meaning. What it does not mean that you need to measure the experimental data by yourself. You need to access the real data either from published data or eventually data that have been obtained from someone else.
I think, even real data doesn't mean it's rigorous. The errors within the experiments, the noises in the data, the special cases... anything that ignored can bury the conclusions that achieved.
Models, whether mathematical or abstract concepts drawn from a real phenomenon, is like a distorting mirror or effects of Photoshop. It is not real, but it is a variation of real world. To same degree, a model can show some conclusions that are highly probably true. If a model can reflect every function and character of a phenomenon in real world. We can trust the evolution result or predictions from the model.
Focus on real data or models are just two ways to think about problems. Of course, combining the two together can make research result more rigorous. But sometimes, one paper only focus on one aspect.
- Badges
- Science topic
- Similar topics
- Books