Given only a limited number of classes, do grad admissions prefer theory or modeling?

#1
I'm currently going into 3rd year, with a messy schedule for the next 4 semesters. Would admissions prefer students to have more modeling courses, or would heavy theory be preferred? Another major problem is that I want to be able to do research in statistics very very soon, and was wondering which ones would be more helpful.

If I go to theory, I would be taking abstract algebra and number theory (interesting classes required for math major), real analysis 1 + 2, and probability theory 1 +2.

I already severely lack actual stat classes though. I am unable to take any statistical computing course, multivariate/regression analysis, time series, data mining, game theory, etc. All I have is a basic probability theory course.

I am aiming to do biostatistics for a phd eventually. I'm worried that I lack too many 'practical' courses that would prohibit me from doing actual research for undergrad, but if I skip theory, I would lose out on tons of mathematical skills. I am very busy and cannot study both at the same time. I also cannot take one theory and one practical course at the same time, for complicated reasons- many factors force me to choose one route or the other.