Can Fold-and-dock protocol applied to membrane proteins

Justin, how did you get on with this? any luck?

Chris

The command above is completely wrong and took me down a blind alley.

My current command looks like this….

/home/chxcja/rosetta3.3_bundles/rosetta_source/bin/minirosetta.linuxgccrelease
-run:protocol broker
-broker:setup setup_init.tpb

-in:file:fasta /home/chxcja/rosetta3.3_bundles/SIT-membrane-fold-dock/input/eliza.FASTA
-in:file:frag3 /home/chxcja/rosetta3.3_bundles/SIT-membrane-fold-dock/input/aat000_03_05.200_v1_3
-in:file:frag9 /home/chxcja/rosetta3.3_bundles/SIT-membrane-fold-dock/input/aat000_09_05.200_v1_3
-in:file:spanfile /home/chxcja/rosetta3.3_bundles/SIT-membrane-fold-dock/input/eliza.span
-in:file:lipofile /home/chxcja/rosetta3.3_bundles/SIT-membrane-fold-dock/input/eliz.lips4
-symmetry:symmetry_definition /home/chxcja/rosetta3.3_bundles/SIT-membrane-fold-dock/input/elizaC4.symm
-symmetry:initialize_rigid_body_dofs
-database /home/chxcja/rosetta3.3_bundles/rosetta_database/

-abinitio:membrane
-membrane:no_interpolate_Mpair -membrane:Menv_penalties -membrane:normal_cycles 100 -membrane:normal_mag 5 -membrane:center_mag 1

-fold_and_dock:rigid_body_cycles 1
-fold_and_dock:rigid_body_frequency 5
-fold_and_dock:rotate_anchor_to_x
-run:reinitialize_mover_for_each_job
-score:weights score13_env_hb
-abinitio:recover_low_in_stages 0
-abinitio:rg_reweight 0.001
-abinitio:use_filters false
-packing:ex1
-packing:ex1:level 1
-packing:ex2
-packing:ex2:level 1
-packing:extrachi_cutoff 0

-evaluation:symmetric_rmsd
-nstruct 20
-out:file:silent_struct_type binary
-relax:quick
-relax:jump_move

-out:path:pdb /home/chxcja/rosetta3.3_bundles/SIT-membrane-fold-dock/pdb
-out:pdb
-out:file:silent silent.out
-no_prof_info_in_silentout
-mute core.io.database
-out:level 500
-out:file:scorefile score.sc

However I’m not sure if this is using any of the membrane settings – i get something that looks vaguely like a pore type structure but the helices are at weird angles (Ie greater than the Menv penalties should allow I think).

I’ll keep people posted on my progress but if anyone has any ideas I’d really like to hear them.

Thanks

Chris

EDITED NOTE: The following command is useless. Do not waste your time with it. I have left it here for the record but please don’t run this….

---

I run fold and dock with the following command (see below) but when I run

/home/chxcja/rosetta3.3_bundles/rosetta_source/bin/score.linuxgccrelease -database ~/rosetta3.3_bundles/rosetta_database/ -in:file:silent default-c4membrane.out -in::file::fullatom -out:output

to get a pdb file of the models the pdbs only contain 1 chain. How can I get the pdb of the complex (in this case a C4 symmetric structure).

Thanks for your help

Chris

***********************************************
/home/chxcja/rosetta3.3_bundles/rosetta_source/bin/membrane_abinitio2.linuxgccrelease
-run:protocol broker
-broker:setup setup_init.tpb

-out:file:scorefile score-membrane.sc
-in:path:database /home/chxcja/rosetta3.3_bundles/rosetta_database/

-file:frag3 aat000_03_05.200_v1_3
-file:frag9 aat000_09_05.200_v1_3

-symmetry:symmetry_definition elizaC4.symm
-out:file:silent default-c4membrane.out
-out:file:silent_struct_type binary
-relax:fast
-relax:jump_move
-symmetry:initialize_rigid_body_dofs

-in:file:fasta eliza.FASTA
-in:file:spanfile eliza.span
-in:file:lipofile eliz.lips4

-abinitio:membrane
-score:find_neighbors_3dgrid
-membrane:no_interpolate_Mpair
-membrane:Menv_penalties
-membrane:normal_cycles 1000
-membrane:normal_mag 5
-membrane:center_mag 1
-nstruct 5

-fold_and_dock:rigid_body_cycles 1
-fold_and_dock:rigid_body_frequency 5
-fold_and_dock:rotate_anchor_to_x

-rg_reweight 0.001
-rigid_body_cycles 1
-abinitio::recover_low_in_stages 0
-rigid_body_frequency 5
-rigid_body_disable_mc
-run:reinitialize_mover_for_each_job
> rosetta.log &

These are the flags that I tried to fold M2 channel, and got a GDTMM 0.70 to the native structure as a tetramer. Hope this helpful –

-run:protocol broker
-broker:setup setup_init.tpb
-database minirosetta_database
-nstruct 100
-out:file:silent_struct_type binary

-in:file:fasta 2kqtA.fasta
-file:frag3 2kqtA.fasta.frags.3mers
-file:frag9 2kqtA.fasta.frags.9mers

-rg_reweight 0.01
-run:reinitialize_mover_for_each_job

-score:find_neighbors_3dgrid

-symmetry:symmetry_definition C4.symm
-symmetry:initialize_rigid_body_dofs
-fold_and_dock::rotate_anchor_to_x

-relax:fast
-relax:jump_move

-abinitio:membrane
-membrane:fixed_membrane
-membrane:no_interpolate_Mpair
-membrane:Menv_penalties

-in:file:spanfile 2kqtA.span
-in:file:lipofile 2kqtA.lips4
-stage2_patch score_membrane_s2.wts_patch
-stage3a_patch score_membrane_s3a.wts_patch
-stage3b_patch score_membrane_s3b.wts_patch
-stage4_patch score_membrane_s4.wts_patch

-membrane:Membed_init
-relax:membrane
-score:weights membrane_highres.wts

-increase_cycles 4.3
-evaluation::gdtmm

Thanks Wangyr, following your suggestions it was clear that there were a few things missing from my commands.

Adding them however goves me an odd result where the actual chains are not in contact with one another.

Can anyone suggest what I might be doing wrong?

Thanks

Chris

Binary output file
http://dl.dropbox.com/u/1184313/SIT-membrane5.out

Commands
############
/home/chxcja/rosetta3.3_bundles/rosetta_source/bin/minirosetta.linuxgccrelease
-run:protocol broker
-broker:setup setup_init.tpb

-in:file:fasta /home/chxcja/rosetta3.3_bundles/SIT-membrane-fold-dock5/input/eliza.FASTA
-in:file:frag3 /home/chxcja/rosetta3.3_bundles/SIT-membrane-fold-dock5/input/aat000_03_05.200_v1_3
-in:file:frag9 /home/chxcja/rosetta3.3_bundles/SIT-membrane-fold-dock5/input/aat000_09_05.200_v1_3
-in:file:spanfile /home/chxcja/rosetta3.3_bundles/SIT-membrane-fold-dock5/input/eliza.span
-in:file:lipofile /home/chxcja/rosetta3.3_bundles/SIT-membrane-fold-dock5/input/eliz.lips4
-symmetry:symmetry_definition /home/chxcja/rosetta3.3_bundles/SIT-membrane-fold-dock3/input/elizaC4.symm
-symmetry:initialize_rigid_body_dofs
-database /home/chxcja/rosetta3.3_bundles/rosetta_database/
-nstruct 20

-abinitio:membrane
-membrane:no_interpolate_Mpair
-membrane:Menv_penalties
-membrane:normal_cycles 200
-membrane:normal_mag 5
-membrane:center_mag 1
-membrane:Membed_init
-abinitio:recover_low_in_stages 0
-abinitio:rg_reweight 0.001

-fold_and_dock:rigid_body_cycles 1
-fold_and_dock:rigid_body_frequency 5
-fold_and_dock:rotate_anchor_to_x
-run:reinitialize_mover_for_each_job
-score:weights score13_env_hb

-packing:ex1
-packing:ex1:level 1
-packing:ex2
-packing:ex2:level 1
-packing:extrachi_cutoff 0
-score:find_neighbors_3dgrid

-evaluation:symmetric_rmsd
-relax:quick
-relax:jump_move

-abinitio:stage2_patch /home/chxcja/rosetta3.3_bundles/SIT-membrane-fold-dock5/input/score_membrane_s2.wts_patch
-abinitio:stage3a_patch /home/chxcja/rosetta3.3_bundles/SIT-membrane-fold-dock5/input/score_membrane_s3a.wts_patch
-abinitio:stage3b_patch /home/chxcja/rosetta3.3_bundles/SIT-membrane-fold-dock5/input/score_membrane_s3b.wts_patch
-abinitio:stage4_patch /home/chxcja/rosetta3.3_bundles/SIT-membrane-fold-dock5/input/score_membrane_s4.wts_patch

-out:pdb
-out:file:silent_struct_type binary
-out:file:silent SIT-membrane5.out
-no_prof_info_in_silentout
-mute core.io.database
-out:file:scorefile score.sc

-relax:membrane
-score:weights membrane_highres.wts
-increase_cycles 4.3

Folding membrane protein in Rosetta is always really tricky. Is this protein a channel that has charged residues in the core? If that’s the case, there are a lot parameterization need to benchmark in energy function for this kind of protein, which no one has done that yet.

Could you run a positive control to see whether this is a target protein problem or a setup problem? Using the M2 channel fasta I posted previously will be an easy and quick test.

Also, it seems like the protein you are working on is really big. Please take a look at Rhiju’s fold’n dock PNAS paper to see what’s the size limitation and how much sampling you need to have.

Best,
Ray

I don’t think that membrane_denovo can handle symmetry, although I don’t know for sure. The error is consistent with this hypothesis.

What does it do if you run it without the membrane stuff? It may be that the membrane scorefunction changes make Rosetta less likely to shove stuff together.