A benchmarking paper (Geistlinger et.al 2020) recommended PADOG as one of the preferred tools for geneset analysis when you have an expression matrix. I am currently implementing this method on our dataset and have encountered a question about the output that I could not find an answer to in the paper, the vignette, or the GitHub issues.

The output contains the following:

1. meanAbsT0: the mean of absolute t-scores
2. padog0: the mean of weighted absolute t-scores

I noticed that some of these values are negative. Given that a t-score can be negative, its absolute value should be positive. If I understood the paper correctly, the weights should also be positive, I am wondering how it is possible to have negative values in the output.

I would greatly appreciate any clarification you could provide on this matter. Thank you in advance for your time and assistance.

Hi Tim, That is a good question. As described in the PADOG paper, there are two normalization steps applied to the two statistics you mention below, and therefore, those normalized pathway scores can be both positive and negative. I will update the padog manual to add this detail. Best, Adi