# errors while running rosetta 3.1

8 posts / 0 new
errors while running rosetta 3.1
#1

Hi
I have two queries regarding ab initio structure prediction with cs-rosetta using rosetta-3.1:
a) For one of the runs while using the script extract_pdb3.com default.out I get the error core.io.silent: ERROR: did not find coordinates for all sequence positions for empty_tag
and only one pdb is generated. The program is still running though.
b) I am running an abinitio structure prediction calculation with rosetta 3.1. I have generated the fragment files from the robetta server. However, on running the command I get the error 'ore.fragments.ConstantLengthFragSet: line too short: Skipping line ' a number of times and the program stops running.
Any help regarding these two problems will be highly appreciated.

Ishita Sengupta.

Post Situation:
Tue, 2011-05-17 18:11
isengupta13

I've never used CSrosetta, but both errors sound to me like you have a number-of-residues mismatch somewhere in your process. For example, you want to predict the structure of a 90 residue protein, but you generated fragments using the FASTA for only 85 residues' worth of protein. Rosetta is not smart enough to silently add things like 6xHis-tags to the ends of proteins (or remove them). That's the first place I'd look for a solution.

Wed, 2011-05-18 08:04
smlewis

I have another question.
The runrosetta3.com script has a few lines which I don't completely understand. I tried looking in the manual, but could not find them.
What do the following terms mean?
-increase_cycles 10 -rsd_wt_helix 0.5 -rsd_wt_loop 0.5 -rg_reweight 0.5 ?

Thanks again,
Ishita Sengupta.

Thu, 2011-05-19 16:21
isengupta13

These are guesses:

core::options::OptionKeys::abinitio::increase_cycles, "Increase number of cycles at each stage of fold_abinitio (or pose_abinitio) by this factor" ).lower(0.001).def(1.0)

core::options::OptionKeys::abinitio::rsd_wt_helix, "Reweight env,pair,cb for helix residues by this factor"
(env, pair, and cb are scorefunction terms in centroid mode. pair handles electrostatics (such as they are); you should be able to look the others up in the Rohl review (2004 methods in enzymology)

core::options::OptionKeys::abinitio::rsd_wt_loop, "Reweight env,pair,cb for loop residues by this factor"

core::options::OptionKeys::abinitio::rg_reweight, "Reweight contribution of radius of gyration to total score by this scale factor"

Mon, 2011-05-23 12:28
smlewis

Thanks a lot!
I found documentation from rosetta 2.3 user guide that explains these terms.
I have another question.
I am running rosetta 3.1 with rdc constraints.
I am not sure how the input file should be or if rosetta 3.1 accepts the dipolar coupling constants at all.
flags file is :

-abinitio:relax
-in:file:fasta trunc.fasta
-in:file:frag3 input/frag3.rosetta.tab
-in:file:frag9 input/frag3.rosetta.tab
-constraints:cst_weight 1
-constraints:cst_file rdc
-out:nstruct 2
-out:file:silent test.out
-out:pdb true
-out:path .

my rdc file is
10 N 10 H -11.407
11 N 11 H -10.639
12 N 12 H -7.626
13 N 13 H -10.212
14 N 14 H -12.215
15 N 15 H -8.948
16 N 16 H -7.531
17 N 17 H -11.159
18 N 18 H -11.65
74 N 74 H -8.485
75 N 75 H -7.471
76 N 76 H -13.03
77 N 77 H -11.264
78 N 78 H -9.004
79 N 79 H -9.42
80 N 80 H -13.559
81 N 81 H -8.236

but everytime i run AbinitioRelax.linuxgccrelease @flags -database

I get the error (last few lines of log file)
core.scoring.constraints: Constraint choice: rdc
core.io.constraints: no section header [ xxx ] found, try reading line-based format... DON'T MIX

ERROR: 10 does not name a known ConstraintType --> check spelling or register new Constraint type in ConstraintFactory!
ERROR:: Exit from: src/core/scoring/constraints/ConstraintFactory.cc line: 58

It seems like the rdc constraint file is not being recognized at all.

In the "using constraints section" of the rosetta-3.1 user manual I could not find any example with rdc data.

Any help would be highly appreciated.

Thanks again,
Ishita.

Thu, 2011-05-26 13:40
isengupta13

The standard constraint file format does not accept RDC data. I've tried in the past to get those who know how to do it to document it, but it hasn't happened yet.

All the RDC stuff is in the 3.2.1 release but I can't trace it back to the actual executeable.

There are multiple options referencing "rdc" in the options list: http://www.rosettacommons.org/manuals/archive/rosetta3.2.1_user_guide/op...

Maybe one of them will work?

Fri, 2011-05-27 08:33
smlewis

Sorry, for bringing this up again but I have the same problem and this seems to be unresolved, yet.

I am also trying to use RDCs in my calculations. I found a reference for this (Sgourakis, 2011, JACS) and was told that this will be available with 3.3.
Now adding the following lines to my 3.3 protocol:
-in:file:rdc rdc_data.dat -score:patch rdc_patch_file

while "rdc_patch_file" is just:
rdc 10.0 atom_pair_constraint 5.0

gives the following message during initialization of the run:
core.scoring: could not parse line in patch-file: rdc 10.0 core.scoring: could not parse line in patch-file: atom_pair_constraint 5.0
The run finishes without problems but I am not sure whether the RDCs were accounted for. Adding arbitrary values as RDCs does not change much, so my guess is, the input-file is ignored.
What may I have missed?

Thu, 2011-09-15 07:03
jurkm

I was too quick posting this question. The patch file has to look like this "rdc = 1.0" or "rdc *= 1.0" with the spaces between name and number. Of course "*=" makes only sense for a predefined score.

Thu, 2011-09-15 08:47
jurkm