Let's consider whether it would be useful to group samples by tissue type, and estimate exposures at the group level, as well as the individual sample level. We could use a scheme where the 'main' estimate is done for the group, and each sample is some perturbation of that estimate. Signatures would be estimated as before.
Let's consider whether it would be useful to group samples by tissue type, and estimate exposures at the group level, as well as the individual sample level. We could use a scheme where the 'main' estimate is done for the group, and each sample is some perturbation of that estimate. Signatures would be estimated as before.