Could anyone please tell me if this makes sense?

We have a cohort with 20 patients, X1, X2, ........, X20. Each patient has a list of somatic mutations. Now, we would like to get profile for each patient in terms of the effect of 30 COSMIC signatures on the patient genome. So if we run YAPSA on the patient subset {X1, X2, X3}, we get a result for the patient 1, XS11. If we run YAPSA on the patient subset {X1, X2, ..., X10}, we get another result for the patient 1, XS12. If we run YAPSA on the patient subset {X1, X2, ..., X20}, we get another result for the patient 1, XS1_3.

Here, XS11, XS12. XS1_3 are all very different.

The effect of COSMIC signature on the patient 1 does not change whatever subset from cohort you choose. With the same YAPSA, different subset selection gives different results.

Is this a problem?