In a letter to Neuron, Drs. Jeong and Liu, from Baylor College of Medicine, called into question the interpretation of the data in a study by Readhead that found elevated levels of HHV-6A and HHV-7 in the brains of Alzheimer’s Disease patients (Readhead 2018). Jeong and Liu state that Readhead et al did “not prove a link between Alzheimer’s disease (AD) and viral load” and that the extremely low expression levels of viral RNA and DNA in these brain samples pose a problem with detection limits. “It may be that such a small signal will never be extractable from these data sets,” they conclude.
Readhead et al said in their response to Neuron that their strategy for mitigating bias was to apply diverse integrative analyses rather than relying on any single finding and that their reported findings collectively link the abundance of many viral species to different facets of Alzheimer’s pathology. They noted that in the field of comparative transcriptomics, it is widely recognized that differences in reasonable analytical strategies can yield substantially different results, with agreement between pairs of methods ranging from 40% to 80%, and that these differences are accentuated with low abundance transcripts, or when differences are subtle, as was the case with their study.
The Readhead group also pointed to new data in a preprint which shows an increased prevalence of HHV-6A in the brains of subjects with Alzheimer’s compared with unaffected control subjects, replicating their published study (BioRXiv).
Letter from Drs. Jeong & LIu: Jeong 2019.
Response from Readhead et al: Readhead 2019