Thanks for your suggestion. I’ll try to increase nrepeat and see if the final model still suggest 3 components. If so, I will end up having 200~300 proteins selected by 3 components. I tried to tune the keeps parameter using multistages you suggested in another post: I started from {20,30,40,50}, and 50 had the lowest BER, then I tried {50,60,70} and 70 had the lowest BER, so I then tried {70,80,90,100}, and 100 had the lowest BER. Every time the model would gave me the upper limit of the choices. I’m currently at 100 and don’t want to go higher. Any suggestions on it?
Another question is that even if I stay with 100, I will end up with ~300 proteins from 3 components. My aim is to look into the biological mechanism, but I’m afraid 300 would still be too many.