Abinitio (3.4) Error: no fragment to compute secondary structure

[Anonymous]

Greetings,

When trying to run AbinitioRelax in Rosetta 3.4, we obtain the following error message:

ERROR: no fragment to compute secondary structure
ERROR:: Exit from: src/core/fragment/SecondaryStructure.cc line: 68

However, the fragment files seem to be read in correctly, based on the following output:

core.init: Mini-Rosetta version unknown from unknown
core.init: command: /cluster/caml/rosetta-3.4/bin/AbinitioRelax.linuxgccrelease -in:file:fasta ./maq3a.fasta -in:file:frag3 ./aat000_03_05.200_v1_3 -in:file:frag9 ./aat000_09_05.200_v1_3 -in:path /cluster/caml/rosetta-3.4/rosetta_database -abinitio:relax -relax:fast -abinitio::increase_cycles 10 -abinitio::rg_reweight 0.5 -abinitio::rsd_wt_helix 0.5 -abinitio::rsd_wt_loop 0.5 -use_filters true -psipred_ss2 ./t000_.psipred_ss2 -kill_hairpins ./t000_.psipred_ss2 -nstruct 1 -out:pdb
core.init: ‘RNG device’ seed mode, using ‘/dev/urandom’, seed=726962078 seed_offset=0 real_seed=726962078
core.init.random: RandomGenerator:init: Normal mode, seed=726962078 RG_type=mt19937
core.init: ROSETTA3_DB not defined
protocols.abinitio.AbrelaxApplication: read fasta sequence: 132 residues
GEIWKQHEDALQKFEEALNQFEDLKQLGGSGSGSGGEIWKQHEDALQKFEEALNQFEDLKQLGGSGSGSGGEIWKQHEDALQKFEEALNQFEDLKQLGGSGSGSGGEIWKQHEDALQKFEEALNQFEDLKQL
protocols.evaluation.ChiWellRmsdEvaluatorCreator: Evaluation Creator active …
core.chemical.ResidueTypeSet: Finished initializing centroid residue type set. Created 1980 residue types
core.io.fragments: reading fragments from file: ./aat000_09_05.200_v1_3 …
core.io.fragments: reading fragment frames from file: ./aat000_09_05.200_v1_3 …
core.io.fragments: read fragments options: top = 25 copies: 1 annotate: no
core.io.fragments: reading fragments from file: ./aat000_03_05.200_v1_3 …
core.io.fragments: reading fragment frames from file: ./aat000_03_05.200_v1_3 …
core.io.fragments: read fragments options: top = 200 copies: 1 annotate: no
core.fragment: compute strand/loop fractions for 0 residues…

The fragments were constructed using the ROBETTA server and the AbinitioRelax protocol is equivalent to that found in the user manual. Any suggestions are appreciated.

Thanks,
Sándor

[Anonymous]

As best I can tell it means Rosetta is generating an empty fragment set. Does the fragment file “look right”, compared to some of the samples in the tutorials/demos/integration tests/your previous experiments?

[Anonymous]

I agree with you that Robetta’s fragments ought to work off-the-bat.

I was surprised to not see the line “rosetta++ fileformat detected! Calling legacy reader…” in your output log, since I’m used to seeing it with Robetta fragments. Indeed, we *should* be seeing that line in the output, as these are old-style fragments, and the trigger for old-style fragments is a first line that says “position:”, which yours does!

Have you tried just re-downloading them from Robetta in case they got corrupted somehow?

[Anonymous]

We’ve just started working with these, so I am unclear if they “look right” or not.
They were created with ROBETTA, so I presume they should be in a compatible format.
The results are available here (for the next few days): http://www.robetta.org/downloads/fragments/30055/

For archival purposes, here are the first few lines of both frag3 and frag9 for review/comments/suggestions.

FRAG3:

position: 1 neighbors: 200

1yoz A 14 G H -63.325 -39.103 176.457 -0.827 4.410 18.545 3 0.000 P 1 F 1
1yoz A 15 E H -55.283 -44.614 -174.673 -0.827 4.410 18.545 3 0.000 P 1 F 1
1yoz A 16 I H -73.140 -44.984 -178.409 -0.827 4.410 18.545 3 0.000 P 1 F 1

1yvu A 75 G L -57.666 -51.893 180.138 -1.146 4.604 52.262 4 0.000 P 1 F 2
1yvu A 76 E L 137.501 -9.529 -179.697 -1.146 4.604 52.262 4 0.000 P 1 F 2
1yvu A 77 I L -73.270 134.182 176.313 -1.146 4.604 52.262 4 0.000 P 1 F 2

2idg A 38 G L 100.918 155.700 -170.522 -0.953 4.882 14.994 5 0.000 P 1 F 3
2idg A 39 E L -71.528 131.406 172.212 -0.953 4.882 14.994 5 0.000 P 1 F 3
2idg A 40 I L -143.394 166.714 -179.436 -0.953 4.882 14.994 5 0.000 P 1 F 3

2icw G 156 G L 48.561 -138.185 -175.204 -0.793 4.643 21.862 3 0.000 P 1 F 4
2icw G 157 E L -108.960 66.209 179.988 -0.793 4.643 21.862 3 0.000 P 1 F 4
2icw G 158 V H -57.981 -41.057 -176.837 -0.793 4.643 21.862 3 0.000 P 1 F 4

FRAG9:

position: 1 neighbors: 200

1aa7 A 22 A H -76.258 -42.976 177.847 0.989 6.761 14.063 3 0.000 P 1 F 1
1aa7 A 23 E H -63.251 -40.524 179.850 0.989 6.761 14.063 3 0.000 P 1 F 1
1aa7 A 24 I H -65.088 -49.041 179.780 0.989 6.761 14.063 3 0.000 P 1 F 1
1aa7 A 25 A H -59.746 -39.075 178.518 0.989 6.761 14.063 3 0.000 P 1 F 1
1aa7 A 26 Q H -63.687 -42.300 179.695 0.989 6.761 14.063 3 0.000 P 1 F 1
1aa7 A 27 R H -64.062 -47.886 178.897 0.989 6.761 14.063 3 0.000 P 1 F 1
1aa7 A 28 L H -56.265 -53.028 -179.461 0.989 6.761 14.063 3 0.000 P 1 F 1
1aa7 A 29 E H -57.562 -37.605 179.825 0.989 6.761 14.063 3 0.000 P 1 F 1
1aa7 A 30 D H -63.867 -48.421 179.342 0.989 6.761 14.063 3 0.000 P 1 F 1

2ezv A 48 G L 60.800 21.849 181.380 2.125 7.069 77.893 4 0.000 P 1 F 2
2ezv A 49 E L -71.412 -31.239 -181.135 2.125 7.069 77.893 4 0.000 P 1 F 2
2ezv A 50 V H -55.760 -24.817 177.877 2.125 7.069 77.893 4 0.000 P 1 F 2
2ezv A 51 I H -61.217 -54.071 -179.633 2.125 7.069 77.893 4 0.000 P 1 F 2
2ezv A 52 V H -70.785 -25.045 179.710 2.125 7.069 77.893 4 0.000 P 1 F 2
2ezv A 53 L H -64.743 -47.860 -179.491 2.125 7.069 77.893 4 0.000 P 1 F 2
2ezv A 54 A H -64.968 -40.699 180.088 2.125 7.069 77.893 4 0.000 P 1 F 2
2ezv A 55 E H -65.738 -48.837 -178.378 2.125 7.069 77.893 4 0.000 P 1 F 2
2ezv A 56 D H -63.911 -45.617 179.759 2.125 7.069 77.893 4 0.000 P 1 F 2

From what I can tell, these look very similar to example fragment files found in the /rosetta-3.4/rosetta_demos/abinitio/input_files directory. The only difference is that the ROBETTA-generated fragment files have additional columns.

Thoughts?

[Anonymous]

Redownloading worked! I also see the line you were referring to in the output:

core.io.fragments: reading fragments from file: ./aat000_09_05.200_v1_3.new …
core.io.fragments: rosetta++ fileformat detected! Calling legacy reader…
core.fragments.ConstantLengthFragSet: finished reading top 25 9mer fragments from file ./aat000_09_05.200_v1_3.new
core.io.fragments: reading fragments from file: ./aat000_03_05.200_v1_3.new …
core.io.fragments: rosetta++ fileformat detected! Calling legacy reader…
core.fragments.ConstantLengthFragSet: finished reading top 200 3mer fragments from file ./aat000_03_05.200_v1_3.new
core.fragment: compute strand/loop fractions for 132 residues…
protocols.abinitio.AbrelaxApplication: run ClassicAbinitio…..
[it loads a bunch of libraries/scorefunctions/etc…]
protocols.jobdist.JobDistributors: Looking for an available job: 1 1 1

However, now I obtain the new error:

ERROR: [ERROR] invalid header input for kill_hairpins file.
ERROR:: Exit from: src/core/scoring/SS_Killhairpins_Info.cc line: 375

When I look at the difference between the ROBETTA output file and the example files, the ROBETTA .psipred_ss2 file is missing the following line (plus one empty line) present in the example file:

# PSIPRED VFORMAT (PSIPRED V2.5 by David Jones)

Including these, however, doesn’t seem to have an effect. The kill_hairpins file is loaded in with the suggested flag:
-kill_hairpins ./t000_.psipred_ss2.new

[Anonymous]

http://www.rosettacommons.org/content/killhairpin-error#comment-4672 (????)

[Anonymous]

Ahhhh! Got it working now! Thanks!

[Author]

Posts