Could logratio methods be useful to analise the generalized propensity score?

I work on statistical methods and applied health data analysis, with interests in compositional data, multivariate count data, clustering, survival analysis, competing risks, and causal inference.