Hi,
I'm trying to cluster structures from a symmetry modeling run (silent file) using the -cluster:input_score_filter flag to exclude structures below a certain energy eg
-cluster:input_score_filter -1780.35
However, I can't seem to get cluster app to pay any attention to the flag, any ideas?
Is there any other way to cluster the top x scoring structures?
Thanks
L
Post Situation:
input_score_filter ignores scores ABOVE a certain energy. Energy units borrow from thermodynamics the idea that lower is better, so high-energy structures are "bad".
check out discussion here :
https://www.rosettacommons.org/content/inputscorefilter-not-taken-progra...
and based on my experience, in 3.5 it is still not integrated.
Let me confuse you more:
In the documentation of 3.5 it is still listed as -input_score_filter
https://www.rosettacommons.org/docs/latest/cluster.html
If you read the help of cluster command: cluster --help, you will not find the option anymore. There is instead one option called -output_score_filter.
If you read the first ca. 50 lines in the output of cluster command, you find neither of such filters listed as possible options. Nevertheless, you can specify them and the program will not complain about unrecognized parameter. Just they are silently ignored in the execution.
Simply put , this score filter option does NOT work. It is not about the energy being higher/lower , beyond/below, better/worse, or whatsoever. It just does not work.