When i run OPLS DA for metabolomics analysis, the performance of my model is very low as shown below. What should i do to fix this issue?

 plasma.oplsda <- opls(dataMatrix, class, predI = 1, orthoI = 2)
OPLS-DA
305 samples x 16391 variables and 1 response
standard scaling of predictors and response(s)
      R2X(cum) R2Y(cum) Q2(cum) RMSEE pre ort pR2Y  pQ2
Total   0.0853    0.865  -0.113 0.184   1   2 0.05 0.15