Check data sources

[ArtPoon]

• verify that sequences are HIV RNA and not DNA
• check that times are in units of days and not weeks, etc.
• other pertinent information (?)

[jpalmer37]

I completed a detailed review of all patients and all research publications to check for the information you requested. Many of the studies published both DNA and RNA sequences, so checking the data required looking at several data sets within each study.

I found that two patient datasets consisted of HIV-1 genomic DNA and thus removed them from the study. All remaining datasets contained RNA.

Although the use of days for timing samples was not explicitly stated in some studies, information on the length of the study, sampling routines, and mentions of days in the results provided reasonable confidence that all datasets use days to measure time.

[ArtPoon]

It's ok to use HIV DNA if those patients are untreated. As long we keep track of which ones are DNA we could also check for an effect.

[jpalmer37]

Oh okay, I wasn't aware that we could do that. I know for a fact that one of the DNA patients underwent no treatment so I'll keep it in the study and make a side note about it. I'll continue searching for treatment information on the other.