Hi,

I have used augustus tool for prediction. I have assembled genome for tulsi. I have used arabidopsis as model organism. i got >80K proteins then i used tobacco and got > 103K. I wonder how could it predict so much when arabidopsis it self has ~40K genes.

i have read some posts related to this, they have suggested doing blast or using clustering methods. But being a naive, Is there a way i can reduce the predicted protein/genes by changing an parameters in .cfg files or how can i make prediction more stringent?

any suggestions are appreciated.