What is the coaccessibility score

[Daniel Gingerich]

I am wondering what the coaccessibility score is in relation to the correlation coefficient (R), R^2, and the p value.  Is the coaccessibility score simply Pearson's correlation coefficient?

I want to look at the correlations between individual pairings of peaks.  Given only the coaccessibility score, how do I know if these peaks are significantly correlated, and also strongly correlated?

Thanks, 

Dan

[hpl...@gmail.com]

Hi Dan,

The coaccessibility score is not quite any of those things... because cicero calculates the inverse covariance scores using graphical lasso on overlapping windows, we needed a way to reconcile the pairs that are calculated twice. We decided to do a mean of the output if the values shared the same sign (both neg or both positive) or NA if the values have different signs.

We don't have a way to assign significance in this context, however the higher the coaccessibility score, the stronger (more confident) the coaccessibility within a dataset (this was backed up by correlating back to various experimental methods in the original paper).

Hope this helps,

Hannah