rosetta scores for mutated structures

[Anonymous]

Hello,
I have written a script for mutating a particular residue to all 20 residue types and scoring all the 20 mutated structures. I am not able to understand how exactly the scoring is done, as some mutations have given unexpected scores.
Following are the lines in command line output. The original residue was LEU at position 27. If you compare the scores of LEU-LEU mutation(score = -4.0111e-07) and LEU-ILE mutation(score = 25.2522), the LEU-ILE score is unexpectedly high.

Is this due to some mistake in protocol or the resfile or any other reason? Please refer the rosetta script and resfile that I have pasted below.

###BEGIN CMD OUTPUT###
ResidueScan: 1qys-mod2_renumbered_0001.resfile 27 A -3.81429
ResidueScan: 1qys-mod2_renumbered_0001.resfile 27 C 8.52039
ResidueScan: 1qys-mod2_renumbered_0001.resfile 27 D 6.09132
ResidueScan: 1qys-mod2_renumbered_0001.resfile 27 E -1.47895
ResidueScan: 1qys-mod2_renumbered_0001.resfile 27 F -4.76405
ResidueScan: 1qys-mod2_renumbered_0001.resfile 27 G 3.35082
ResidueScan: 1qys-mod2_renumbered_0001.resfile 27 H -12.9753
ResidueScan: 1qys-mod2_renumbered_0001.resfile 27 I 25.2522
ResidueScan: 1qys-mod2_renumbered_0001.resfile 27 K 2.87281
ResidueScan: 1qys-mod2_renumbered_0001.resfile 27 L -4.0111e-07
ResidueScan: 1qys-mod2_renumbered_0001.resfile 27 M -9.61207
ResidueScan: 1qys-mod2_renumbered_0001.resfile 27 N -1.00938
ResidueScan: 1qys-mod2_renumbered_0001.resfile 27 P 11.7584
ResidueScan: 1qys-mod2_renumbered_0001.resfile 27 Q -1.14136
ResidueScan: 1qys-mod2_renumbered_0001.resfile 27 R 4.13615
ResidueScan: 1qys-mod2_renumbered_0001.resfile 27 S -8.53399
ResidueScan: 1qys-mod2_renumbered_0001.resfile 27 T -6.68773
ResidueScan: 1qys-mod2_renumbered_0001.resfile 27 V -4.50213
ResidueScan: 1qys-mod2_renumbered_0001.resfile 27 W 4.70038
ResidueScan: 1qys-mod2_renumbered_0001.resfile 27 Y -13.7066
###END CMD OUTPUT###

###BEGIN SCRIPT###
Please see attached file viz. residueScan.txt
###END SCRIPT###

###BEGIN RESFILE###
nataa
start
27 A ALLAAwc
###END RESFILE###

[Anonymous]

At the moment I can’t see anything obviously wrong with your script.

One thing to keep in mind is that the FilterScan, like most things in Rosetta, is stochastic. If this is the result of a single run there’s a chance you got a bad run, and your isoleucine got stuck in a bad position. Is it giving you a consistently bad delta if you run it multiple times, or is the high number just a one-off?

Ideally, you’d be able to look at the structure of the isoleucine variant it’s creating, so that if there is an obvious issue, you can investigate that. A “dump pdb” functionality has been added to FilterScan in the development branch, but unfortunately it isn’t in 3.3. A work around is to “fake” what FilterScan is doing with another script. To create the mutant, FilterScan effectively creates a resfile forcing the amino acid identity at the position, then applies a ReadResfile and DesignAround operation (with design_shell=8.0, repack_on=0 – in combination with the resfile this shouldn’t change amino acid identities except for the mutant), and then uses them in a PackRotamers mover. It follows this up by applying the relax mover (the MinMover in your case).

[Author]

Posts