A group of scientists from the Baylor School of Medicine analyze data from previous studies and deny the relationship between the two ailments

American researchers have presented a study that denies that there is a relationship between Alzheimer’s and herpes virus, as previous studies had suggested and proposed a new methodology for the analysis of large-scale sequencing data.

A theory that has gained weight in the last year is that certain microbial infections such as those caused by viruses can trigger the disease of Alzheimer. In a 2018 study, reference was made to the fact that there were higher levels of human herpesvirus 6A (HHV-6A) and 7 (HHV-7) in postmortem brain tissues of more than 1,000 patients with Alzheimer’s disease. Much more than those found in brain tissues of subjects of healthy ageing or those who suffered a different neurodegenerative disease.

However, Hyun-Hwan Jeong, one of the co-authors of this new study, and several colleagues analyzed datasets from the 2018 study, using identical statistical methods with rigorous filtering and four commonly used statistical tools, were unable to obtain the same results.

Number problem

The team was motivated to re-analyze the data from the previous study because they observed that while the p-values (a statistical parameter that predicts the probability of obtaining the observed results of a test, assuming other conditions are correct) were very significant, they were attributing to data in which the differences were not visually appreciable.

In addition, p-values ​​did not fit with simple logistic regression, a statistical analysis that predicts the result of the data as one of the two defined states. In fact, after several types of rigorous statistical tests, they found no link between the abundance of herpes or RNA viral DNA and the likelihood of Alzheimer’s disease in the analyzed group.

“Our study highlights one of the possible pitfalls of excessive dependence on p-values. While p-values ​​are a very valuable statistical parameter, they cannot be used as an independent measure of statistical correlation: procedural data sets High performance should still be carefully plotted to visualize the dissemination of the data,” he tells Neuroscience News. “The data sets must also be used in conjunction with accurately calculated p-values ​​to make genetic disease associations statistically correct and biologically significant.”

New analysis framework

“Our goal in pursuing and publishing this study has been to generate tools and guidelines for the analysis of “big data“ so that the scientific community can identify treatment strategies that are likely to benefit patients,” says Zhandong Liu, another author of the study, which have been published in the scientific journal ‘Neuron’.