Can experiments detect differences that matter?
When conducting experiments, it is important to understand if they are able to detect differences that matter. In this section opening to the first question, the author begins discussing how profiling a microbiome leads to categorization at the level of phyla, species or genes. This can be an issue as it only allows for coarse sorting as the criterion used by researchers is the different ratio of bacteria to distinguish microbes. An example they applied this to in the article was if an experiment was used to characterize animal communities, based on the ratio between 100 birds and 25 snails and the ratio between 8 fish and 2 squids. In this instance, both ratios would appear identical. Even in a single species strain can still differ in their genes. Nowadays modern technology has made it possible to study more genes in a sample and decipher metabolic networks which can let us know the biochemical reactions a microbiome can process. Ultimately this is crucial as it can lead to the identification of gene combinations. No matter what, it is difficult to completely know an outcome to an entity unless the networks are probably characterized. The ability to identify functional differences in related genes also plays an important role to understand the genes or networks. Genomes are poorly understood but are also important as they can make crucial differences in what metabolic networks do. Overall an experiment has the opportunity to detect differences or not play a significant role at all but we need to be able to identify functional differences in genes that are closely related from a sequence first.
Does the study show causation or correlation?
Moving on to question 2, the author goes on to discuss how an experiment was conducted in two directions, inverse and reverse directions in order to know if the study showed causation or just correlation. This is very important to question when interpreting scientific literature. An example given mentioned a study about gut microbiomes and diet in an article from 2012 which proposed a causal relationship after conducting a study between the gut microbiomes of old people living in care homes and old people living in the community. Although the data and proposal were fit together the author stated that the reverse causality and the potential for poor health to alter the gut microbiome was not investigated. The less active immune system and differences in the digestion of frailer people could have led to changes in the microbiome. In this case, the conclusion about the causal relationship was incorrect. It is mentioned this confusion occurs often. Therefore, I believe that understanding if a study shows causation or correlation is important as it aids researchers to have an overview and comprehend the relationship between the factors at hand.
What is the mechanism?
It is important to understand what the mechanism is as it helps researchers to have a better understanding of what components of the microbiome they are studying. They discuss how scientists are always taught catechism and the relationship between causation and correlation. This helps comprehend a correlation which can contain a causal relationship but researchers do not know what makes the relationships. Recent studies have been able to identify functional elements, different taxa, and specific characteristics. It can also define actions of elements in the microbiome related to the biochemical activity. Although a reductionist approach is needed to pinpoint if and how microbiome affects human health. This contributes to an overall improved comprehension of the study.
How much do experiments reflect reality?
Moving on to the fourth question, it is important to understand how much experiments reflect reality. Even if the microbiome implements an effect this may not be relevant to the cause of symptoms. For example, the article discusses the study about gut flora and weight gain which multiple studies have found associations between. The experiments were conducted on germ-free mice which did not represent the natural state of animals and did not have healthy owing. Therefore, the study did not include the responses in animals with flouring microbiomes related to different adaption between mice and their microbiomes and humans. The question at hand provides researchers more insight about their subjects and also evaluates if the results from the study are useful and relevant.
Could anything else explain the results?
With any study there are so many variables at hand that can have an impact on the results. It is important to control as many of those variables as possible but not all of them are able to be. This question addresses if anything such as those variables could explain the results other than the ones intended to. This is important for researchers to take into consideration so they can attempt to control them and attempt to understand if they are altering the study. These variables could affect the results of their study and the way to analyze the data, generate the hypotheses and evaluate the conclusion. For example, bacteria affects humans but whether or not there are possible factors contributing to these effects. It is important to know about contributing factors in a study in order to make sure the results are not affected by these factors before coming to a conclusion and assuming what was being tested was the only cause.
Which is most helpful when discussing controversy and why?
Overall I feel as though the most helpful question when discussing controversy is question one, “Can experiments detect differences that matter?”. If the experiments were not able to convey relevant changes or new information it would be hard to come to any conclusion or have any chance to argue a controversy. This is the most important question when interpreting literature.
Recent Comments