I am running a panel cointegration test using the Pedroni approach. The panel consists of 53 countries across Africa for 19 years. My dependent variable(foreign direct investment) is stationary at levels(according to the LLC results), together with two other explanatory variables. The remaining three explanatory variables are stationary at first difference.

Is it still possible to run a panel cointegration test?
Note:
According to Statav16.1 manual( page 74), the example given for the dependent variable is I(1). I am wondering whether it is still possible to examine long-run convergence when the dependent variable is stationary at level but the Independent variable only becomes stationary at 1st difference.