LM curve shows the relationship among currency circulation (M), price level (P), interest rate level (R) and output (GDP). The choice of explanatory variables and explained variables in econometric model depends on the purpose of your model.

If we want to explain the influence of money circulation (M) on national output (GDP), it is obviously necessary to set GDP as the explained variable, that is, the Y variable.

If we want to explain the influence of national output on monetary policy, we should set m as the explained variable.

If your model has only two variables, GDP and M, then I suggest that you don't simply decide who is the explanatory variable and who is the explained variable, because both are time series data, so it involves co-integration model, so you should adopt EG two-step method or other time series models instead of simply making regression. Otherwise, there will be false regression.