Program : SOLEX
Version : 990126
Author : Gerard J. Kleywegt & T. Alwyn Jones,
Dept. of Cell and Molecular Biology,
Uppsala University, Biomedical Centre, Box 590,
SE-751 24 Uppsala, SWEDEN
E-mail : gerard@xray.bmc.uu.se
Purpose : extraction of top solutions from ESSENS runs
Package : RAVE
Reference(s) for this program:
* 1 * T.A. Jones (1992). A, yaap, asap, @#*? A set of averaging programs. In "Molecular Replacement", edited by E.J. Dodson, S. Gover and W. Wolf. SERC Daresbury Laboratory, Warrington, pp. 91-105.
* 2 * G.J. Kleywegt & T.A. Jones (1994). Halloween ... Masks and Bones. In "From First Map to Final Model", edited by S. Bailey, R. Hubbard and D. Waller. SERC Daresbury Laboratory, Warrington, pp. 59-66.
* 3 * G.J. Kleywegt & T.A. Jones (1997). Taking the fun out of map interpretation. CCP4/ESF-EACBM Newsletter on Protein Crystallography 33, January 1997, pp. 19-21. [http://alpha2.bmc.uu.se/usf/factory_7.html]
* 4 * G.J. Kleywegt & T.A. Jones (1997). Template convolution to enhance or detect structural features in macromolecular electron-density maps. Acta Cryst D53, 179-185. [http://www.iucr.ac.uk/journals/acta/tocs/actad/1997/actad5302.html]
* 5 * G.J. Kleywegt & R.J. Read (1997). Not your average density. Structure 5, 1557-1569. [http://www4.ncbi.nlm.nih.gov/htbin-post/Entrez/query?uid=9438862&form=6&db=m&Dopt=r]
* 6 * G.J. Kleywegt & T.A. Jones (2037 ?). Convenient single and multiple crystal real-space averaging. To be published ???
* 7 * G.J. Kleywegt & T.A. Jones (1999 ?). Chapter 25.2.6. O and associated programs. Int. Tables for Crystallography, Volume F. To be published.
961120 - 0.1 - first version
961228 - 1.0 - first production version
990126 - 1.0.1- if rotation file too big, tell the user how much
memory should be allocated to it (MASKSIZE)
NOTE: This program is sensitive to the environment variable CCP4_OPEN. If this variable has *not* been set, you will not be able to create any CCP4 maps. If this happens, the program will abort execution on startup. To fix this, put the following line in your .login or .cshrc file: setenv CCP4_OPEN UNKNOWN
SOLEX allocates memory for maps and masks dynamically. This means that you can increase the size of maps and masks that the program can handle on the fly:
1 - through the environment variables MAPSIZE and MASKSIZE (must be in capital letters !), for example put the following in your .cshrc file:
----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- setenv MAPSIZE 8000000 setenv MASKSIZE 3000000 ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE -----
2 - by using command-line arguments MAPSIZE and MASKSIZE (need not be in capitals), for example:
----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- run solex mapsize 10000000 masksize 5000000 |& tee solex.out ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE -----
Note that command-line arguments take precedence over environment variables. So you can set the environment variables in your .cshrc file to "typical" values, and if you have to deal with a map and/or mask which is bigger than that, you can use the command-line argument(s).
If sufficient memory cannot be allocated, the program will print a message and quit. In that case, increase the amount of virtual memory (this will not help, of course, if you try to allocate more memory than can be addressed by your machine (for 32-bit machines, something 2**32-1 bytes, I think), or reduce the size requirements.
SOLEX needs space for 1 map and 1 mask.
In the original version of ESSENS, the program could save the best- fitting orientation of a search fragment or molecule to a PDB file, but extracting the second-best solution etc. was tedious to say the least. In addition, in feature extraction mode, all you got was a map showing where helices or strands might fit, but the information regarding the best orientation was lost.
From version 2.0 onwards, ESSENS saves two maps (a "score map" and a "display map"), as well as a "rotation file" which contains the Euler or polar angles which gave the best fit of the template to your electron density.
This new program SOLEX (SOLution EXtractor) uses these files to extract the best solution(s). For alpha-helices and beta-strands it can even attempt to auto-trace (but this is still an experimental piece of code !!!).
----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- *** SOLEX *** SOLEX *** SOLEX *** SOLEX *** SOLEX *** SOLEX *** SOLEX ***Version - 961228/1.0 (C) 1992-97 Gerard J. Kleywegt & T. Alwyn Jones, BMC, Uppsala (S) User I/O - routines courtesy of Rolf Boelens, Univ. of Utrecht (NL) Others - T.A. Jones, G. Bricogne, Rams, W.A. Hendrickson Others - W. Kabsch, CCP4, PROTEIN, E. Dodson, etc. etc.
Started - Sat Dec 28 22:58:22 1996 User - gerard Mode - interactive Host - sarek ProcID - 28664 Tty - /dev/ttyq25
*** SOLEX *** SOLEX *** SOLEX *** SOLEX *** SOLEX *** SOLEX *** SOLEX ***
Reference(s) for this program:
* 1 * G.J. Kleywegt, Uppsala University, Uppsala, Sweden, Unpublished program.
For more/related/up-to-date/complete references, check: http://alpha2.bmc.uu.se/usf/references.html
For the most recent version of the manual, check: http://alpha2.bmc.uu.se/usf/gerard_manuals.html
*** SOLEX *** SOLEX *** SOLEX *** SOLEX *** SOLEX *** SOLEX *** SOLEX ***
Allocate maps of size : ( 10000000) Allocate masks of size : ( 5000000) Max nr of atoms : ( 10000) ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE -----
Give the name of your "score map" (*not* the "display map" !) from a previous ESSENS run.
----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- Input ESSENS score map file ? ( ) score_b20.E Input ESSENS score map file : (score_b20.E) Read header Input map : (score_b20.E) FORMATTED OLD file opened on unit 11 Logical name: score_b20.E, Full name: score_b20.E(Q)QOPEN allocated # 1 User: gerard Logical Name: score_b20.E Status: READONLY Filename: score_b20.E
File name for input map file on unit 11 : score_b20.E file size = 3276224 ; logical name score_b20.E
Number of columns, rows, sections ............... 100 89 92 Map mode ........................................ 2 Start and stop points on columns, rows, sections -12 87 11 99 23 114
Grid sampling on x, y, z ........................ 100 110 64 Cell dimensions ................................. 91.80000 99.50000 56.50000 90.00000 90.00000 90.00000 Fast, medium, slow axes ......................... Z X Y Minimum density ................................. -5.83801 Maximum density ................................. 23.92435 Mean density .................................... -5.58603 Rms deviation from mean density ................. 1.53606 Space-group ..................................... 19 Number of titles ................................ 4
Titles : new 2Fo-Fc map Created by AVE V. 931216/3.0 at Wed Dec 22 18:45:45 1993 for user gerard Created by AVE V. 950118/3.1 at Mon Jul 17 19:04:35 1995 for user gerard Created by ESSENS V. 961120/2.0 at Thu Nov 21 23:04:40 1996 for user gerard
Parameters as read from the map file Origin ...................... 11 23 -12 Extent ...................... 89 92 100 Grid ........................ 100 110 64 Cell axes ................... 91.80 99.50 56.50 Cell angles ................. 90.00 90.00 90.00 UVW (fast, medium, slow) .... Z X Y
Header done Map read OK Closing BINARY CCP4 map on unit : ( 11) Cell axes (A) : ( 91.800 99.500 56.500) Cell angles (d) : ( 90.000 90.000 90.000) Grid axes (pts) : ( 100 110 64) Origin (pts) : ( 11 23 -12) Extent (pts) : ( 89 92 100) Grid spacing (A): ( 0.918 0.905 0.883) ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE -----
Supply the name of the rotation file (produced in a previous run of ESSENS).
----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- Input ESSENS rotation file ? ( ) b20.rot Input ESSENS rotation file : (b20.rot) Reading rotation file ... Cell axes (A) : ( 91.800 99.500 56.500) Cell angles (d) : ( 90.000 90.000 90.000) Grid axes (pts) : ( 100 110 64) Origin (pts) : ( 11 23 -12) Extent (pts) : ( 89 92 100) ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE -----
Provide the name of the PDB file containing your template or (partial) search model. This must be *exactly* the same PDB file as the one you used in the ESSENS calculations !
----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- Name of template PDB file ? (alpha.pdb) /home/gerard/progs/average/essens_beta.pdb Name of template PDB file : (/home/gerard/progs/average/essens_beta.pdb) Number of atoms : ( 25) Nr of H stripped : ( 0) Centre of gravity of atoms : ( 0.000 0.000 0.000) Using central CA atom as pivot point Atom : (ATOM 12 CA ALA 3 -0.650 -0.185 -0.410 1.00 20.00) Pivot point : ( -0.650 -0.185 -0.410) Furthest atom (A): ( 7.883) ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE -----
Select one of the four possible solution extraction modes. Note that the auto-trace facility is very experimental !
----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- SOLEX can extract top solutions in one of 3 modes: 1 = each to a separate file (e.g., domain) 2 = all to one file (e.g., ligand) 3 = try to trace alpha-helices 4 = try to trace beta-strands Solution extraction mode ? ( 1) 4 Solution extraction mode : ( 4) ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE -----
For extraction mode 1, provide the basename of the output PDB files (e.g., a basename "essens" will give you PDB files which are called "essens1.pdb", "essens2.pdb", etc.).
For the other extraction modes, provide the name of the new PDB file.
----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- Name for output PDB file ? (solex.pdb) btrace.pdb Name for output PDB file : (btrace.pdb) ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE -----
For extraction modes 3 and 4 only, a DEJAVU SSE file will be created. This can be used as input to DEJAVU (operating in "Bones search" mode) to look for any other proteins in the PDB which have a similar set of strands/helices. If you have run ESSENS/SOLEX for both helices and strands, merge the two SSE files before running DEJAVU. See the DEJAVU manual for more details.
For extraction modes 3 and 4, you may have renamed all but one of your CA atoms (e.g., to " CAX") in order to force ESSENS to use a particular CA atoms as the pivot point. SOLEX will rename such atoms to " CA ".
----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- Name for output DEJAVU SSE file ? (beta.sse) beta.sse Name for output DEJAVU SSE file : (beta.sse)Renaming " CA?" atoms to " CA " ... Removing all but " CA " atoms ... Nr of remaining atoms : ( 5) ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE -----
Provide the lowest value for acceptable/interesting solutions. You obtain this number for helices/strands by contouring the ESSENS "display map" at a level where you can see the strands/helices clearly, and not too much (connected) noise.
----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- Cut-off score for top solutions ? ( 8.908E+00) 13.0 Cut-off score for top solutions : ( 1.300E+01) ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE -----
Provide the maximum number of solutions you want SOLEX to generate.
----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- Max nr of top solutions ? ( 25) 50 Max nr of top solutions : ( 50) ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE -----
The output below is for a run in which beta-strands are extracted, and where SOLEX tries to trace them.
First, the top solutions are located:
----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- Extracting solutions ... Sol # 1 Value 2.3924E+01 @ 32 49 49 Rot 340 100 140 Sol # 2 Value 1.8874E+01 @ 44 40 58 Rot 100 60 20 Sol # 3 Value 1.8343E+01 @ 44 41 52 Rot 260 60 100 ... Sol # 50 Value 1.3056E+01 @ 37 51 65 Rot 200 60 120 Max number of solutions done ... ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE -----
Subsequently, the auto-tracing routine kicks in:
----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- Nr of CA atoms generated : ( 250) Trying to connect them ...Starting TRACER ... Nr of solutions : ( 50) Max allowed : ( 500)
Calculate vectors, angles and distances ... Try to extend bits ... Prepend 4 at N-terminus of 1
Calculate vectors, angles and distances ... Try to extend bits ... Invert 5 for use with 1 Append 5 at C-terminus of 1
Calculate vectors, angles and distances ... Try to extend bits ... Invert 10 for use with 1 Prepend 10 at N-terminus of 1
...
Calculate vectors, angles and distances ... Try to extend bits ...
Could not extend any further ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE -----
This part of the program tries to merge bits of strand/helix that are parallel and lie close in space. Only the CA atoms are retained !
When it is finished, all stretches which have been extended by at least one CA atom are written to the PDB file (with residues renumbered to 1, 2, 3, ...). The occupancy of each CA atom is replaced by the number of the solution (1, 2, ...), and the B-factor by the score of the solution in the "score map" (i.e., higher values, better fits). Simultaneously, an SSE file for DEJAVU is created.
For PDB files created by extraction mode 1 and 2, the occupancy is again the number of the solution (but divided by 100), and the B-factor is set to: score - average_score / standard_deviation_score (i.e., higher numbers mean better fits).
----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- Nr of stretches left : ( 19) Stretch 1 has 12 atoms (solution 2 ) Stretch 2 has 9 atoms (solution 1 ) Stretch 3 has 8 atoms (solution 7 ) Stretch 4 has 8 atoms (solution 9 ) Stretch 5 has 8 atoms (solution 23 ) Stretch 6 has 7 atoms (solution 3 ) Stretch 7 has 7 atoms (solution 12 ) Stretch 8 has 6 atoms (solution 13 ) Written PDB file Written DEJAVU SSE fileCPU total/user/sys : 6.6 6.3 0.3
*** SOLEX *** SOLEX *** SOLEX *** SOLEX *** SOLEX *** SOLEX *** SOLEX ***
Version - 961228/1.0 Started - Sat Dec 28 23:15:34 1996 Stopped - Sat Dec 28 23:15:56 1996
CPU-time taken : User - 6.3 Sys - 0.3 Total - 6.6
*** SOLEX *** SOLEX *** SOLEX *** SOLEX *** SOLEX *** SOLEX *** SOLEX ***
>>>>>> This program: (C) 1992-97, GJ Kleywegt & TA Jones <<<<<< >>>> E-mail: gerard@xray.bmc.uu.se or alwyn@xray.bmc.uu.se <<<< > http://alpha2.bmc.uu.se/usf/gerard_manuals.html <
*** SOLEX *** SOLEX *** SOLEX *** SOLEX *** SOLEX *** SOLEX *** SOLEX ***
STOP ... Toodle pip ... statement executed ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE -----
Note that this operation hardly takes any CPU time ! Of course, the results are not necessarily fantastic. We have tested it only with 20-degree rotation angles so far and what happens is (in the case of P2 myelin) is that SOLEX finds large bits of many strands, but not all of them run in the correct direction. So: beware !
For extraction mode 1, read each of the top solutions into O, and judge which one (if any) fits your map best.
For extraction mode 2 (e.g., for ligands, Trp sidechains, etc.), read the PDB file into O, draw the solutions and decide which ones to keep.
For extraction mode 3 and 4, read the PDB file into O, e.g.:
----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- s_a_i btrace.pdb btra mol btra paint_ramp atom_b ; red blue ca ; end bell ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE -----
If you use this way to paint_ramp the molecule, then: the more green/blue the CA atoms, the better the fit.
Of course, you can use the SSE file to do a DEJAVU search (use the "Bones search" option).
You may want to look up the following two references in which the use of DEJAVU to work at the level of rough Bones-derived secondary- structure elements was first described:
(1) Kleywegt, G.J. and Jones, T.A. (1994). Halloween ... masks and bones. In "From First Map to Final Model", edited by S. Bailey, R. Hubbard and D. Waller. SERC Daresbury Laboratory, Warrington, pp. 59-66.
(2) Kleywegt, G.J. and Jones, T.A. (1997 ?). Detecting folding motifs and similarities in protein structures. Meth Enzymol ?, ?-?.
An SSE file created by SOLEX may look as follows (if you have run ESSENS with both strands and helices, you can merge their SSE files of course):
----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- ! Created by SOLEX V. 961228/1.0 at Sat Dec 28 23:36:51 1996 for user gerard ! MOL bone NOTE auto-generated by SOLEX PDB btrace.pdb ! BETA 'B1' ' 1' ' 12' 12 61.43 60.73 47.76 33.97 55.75 27.06 BETA 'B2' ' 13' ' 21' 9 44.24 63.08 16.44 37.40 64.56 41.58 BETA 'B3' ' 22' ' 29' 8 56.31 63.65 17.51 44.11 72.87 32.13 BETA 'B4' ' 30' ' 37' 8 49.36 51.47 27.01 61.21 66.47 37.90 BETA 'B5' ' 38' ' 45' 8 57.25 53.27 22.42 59.65 74.87 31.87 BETA 'B6' ' 46' ' 52' 7 45.76 52.50 31.42 59.24 63.58 40.97 BETA 'B7' ' 53' ' 59' 7 62.51 73.28 34.79 52.24 58.42 26.17 BETA 'B8' ' 60' ' 65' 6 47.19 65.18 19.62 39.41 67.92 33.35 ENDMOL ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE -----
These 8 beta-strands were extracted from the averaged P2 myelin map. Running them through DEJAVU with the suggested "Bones search" settings gives one hit if we require that at least 7 strands must match:
----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- Entry 380 = 1ftp = fatty-acid-binding protein - desert locust (sch ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE -----
This is of course a perfect hit, since FABP and P2 myelin have the same fold ! If we relax the search and require only 6 strands to match, we get:
----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- ===> Option ? (INCR) incr********** NEW QUERY **********
Elements : ( B1 B2 B3 B4 B5 B6 B7 B8) Min nr of residues for SSEs ? ( 5) ................... ( B1 B2 B3 B4 B5 B6 B7 B8) Min nr of elements to match (0 = abort) ? ( 7) 6
Is this a BONES search ? (Y) BONES search mode
BONES search; will do lsq_centroid BONES suggested values: 1 5 10.0 5.0 0.2 Define how much the nr of residues in SSEs may differ by defining how many residues shorter or longer SSEs in the database may be compared to those in your protein. Max nr of residues "too short" ? ( 1) Max nr of residues "too long" ? ( 5) Mismatch element length ? ( 10.000) Mismatch distances ? ( 5.000) Mismatch cosines ? ( 0.200) Weights for nr res, length, dist, cos, rmsd BONES suggested values: 0 0 10 10 10 Weights for scoring ? ( 0.010 0.010 0.327 0.327 0.327) Normalised weights : ( 0.010 0.010 0.327 0.327 0.327)
Possible distance criteria: C => centre-to-centre H => MIN head-tail and tail-head (anti-parallel) T => MIN head-head and tail-tail (parallel) I => MIN of all these distances A => MAX of all these distances BONES suggested value: C !!! Which distances (C/H/T/I/A) ? (C) Extensive output ? (N) BONES suggested value: NO !!! Conserve directionality ? (N) BONES suggested value: Y Conserve absolute motif ? (Y) BONES suggested value: NO !!! Conserve neighbours ? (N) Attempt to avoid multi-chain hits ? (Y) Attempt to avoid identical proteins ? (Y) Create O macro file ? (Y) O macro file ? (beta_lsq.omac) BONES search; cannot use LSQMAN
Nr of elements recognised in query : ( 8) Indices : ( 1 2 3 4 5 6 7 8) Nr of elements of each type : ( 0 8)
********** 1cbs ********** 152 ********** [cellular retinoic-acid-binding protein type ii co - human (homo sapie ] [/nfs/scr_uu5/gerard/dejavu/1cbs.pdb ]
...
Nr of database entries : ( 1381) Nr of selected entries : ( 1381) Nr of matching entries : ( 4) Nr of hits (total) : ( 12)
Entry 152 = 1cbs = cellular retinoic-acid-binding protein type ii co - human (homo sapie Entry 219 = 1crb = cellular retinol binding protein (crbp) complexed - rat (rattus rattu Entry 380 = 1ftp = fatty-acid-binding protein - desert locust (sch Entry 825 = 1pmp = p2 myelin protein (p2) - bovine (bos taurus
2 CPU total/user/sys : 1.6 1.6 0.0 ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE ----- EXAMPLE -----
All four hits are correct ! So now all we have to do is to execute the
O macro produced by DEJAVU, select one of the four PDB entries as our
starting model, use Mutate_replace a bit, rebuild the model a bit, and
throw it into X-PLOR !
(Actually, the orientation of 1CRB is not the correct one, which would
probably have been avoided if the helices had been used as well. Also,
the alignment is very rough, so some rigid-body refinement would be
necessary in a real-life case !)
None, at present.
Created at Sat Jan 30 02:51:11 1999
by MAN2HTML version 971024/1.6