ESCET Version 0.7 User Manual
Contents
1. aset_read ifile 1TAT pdb sele occ gt 0 0 aset_egen npar 25408 nobs 18194 cp1 97 0 dmin 3 0 rfree 30 0 aset_select tset 5 sele chainid A run difference distance matrix analysis ddm setl 1 5 esd_scaled on lolim 2 0 hilim 5 0 ssplot on rbplot on rb_iterate 3 stop This script produces the plot shown in Figure 3 4 14 5 00 1 5 0 150 5 00 2 0 0 3A 03 20 Figure 3 4 Difference distance matrix between the structures of tryptophan synthase in complex with F IPP TRPS and in complex with F IPP and amino acrylate TRPS P In the lower left half the error scaled difference distance matrix is displayed using upper and lower cutoffs of 1 5 and 5 00 respectively In the upper right half the ordinary difference matrix is displayed using a lower cutoff of 0 3 A corresponding to approximately the lo level as determined by the o4 method For scaling an upper cutoff of 2 0 A has been employed Both matrices underwent 3x3 binning prior to being displayed Below the matrix secondary structure elements are shown schematically open boxes stand for helices filled boxes for beta sheets 15 3 4 Mersacidin Mersacidin is a polypeptide antibiotic containing 20 amino acids that crystallizes with six molecules in the asymmetric unit The structure was solved and refined against merohedrally twinned data to 1 06 resolution as described in 6 The conformations of the six molecules2. DP 5 helix pi helix S bend I hydrogen bonded turn e extension of beta strand g extension of 3 10 helix h extension of alpha helix 4 used in PROCHECK but I don t know what it means no secondary structure Table 2 1 Identifiers used for secondary structure elements If a pdb file is read the program will try to figure out the secondary structure assignment from the HELIX and SHEET records If you are reading the data from a non pdb file a pdb file without the HELIX and SHEET records or if you are not happy with the assignment read you can modify the atom property ssid using the aset_amod command The two commands aset_amod set 1 ssid H sele resi in 3 20 34 40 aset_amod set 1 ssid E sele resi in 24 30 44 50 will modify the atoms in set 1 such that residues 3 to 20 and 34 to 40 will become qa helical and residues 24 to 30 and 44 to 50 will become sheet Chapter 3 Identification of the rigid part of a macromolecule Please try to roughly understand the paper on the sub ject 7 before trying to run the program As it stands now the analysis of a set of conformers will proceed as follows Read a bunch of models from one in the case of NCS related molecules or several coor dinate files aset_read command If no standard uncertainties have been read from the coordinate file as is the case most of the time unless you are in the fo
3. e In previous version the Cruickshank error estimate was calculated deriving some of the input numbers from the current atom set If the calculation was done after removing some atoms or only retaining the CA s this could give wrong results The program now remembers the numbers from the original pdb file and uses those as default values e aset_egen now has a keyword esd_blim to select the lowest accepts B value e if secondary structure is plotted now the data for this come from a reference set The number of this set defaults to the first set selected but can be changed using the setr keyword of ddm e sele now works properly with chainid e aset_ write can now write selected sets of atoms e The program can now hold 100 atom sets e Some cleanup on eps output files e Started a chapter on tips amp tricks Version 0 1e 9 Jan 2001 e Chorismate Mutase was included as an example into the user s manual e Hopefully the use of Equation 26 or 27 from Cruickshank s paper 2 is now somewhat clearer e When reading coordinates from a file using the aset_read command atoms can now be selected directly i e if you only want CA atoms from a file called fname pdb use aset_read ifile fname pdb sele name CA 39 Version 6 0 1d 14 Nov 2000 A number of things concerning the refinement are now extracted from the pdb file and passed on to the aset_egen command I am not sure how to determine a sensible number of p
4. 37 5394 5406 1998 41
5. ESCET produces a POSTSCRIPT file that only contains the error scaled difference distance matrix Sometimes it may be useful to have plot containing both the standard and the error scaled difference distance matrix This can be achieved by replacing the ddm command in ribo1 inp with the following lines the complete script can be found in the examples directory as ribo2 inp ESCET input file ribo2 inp compare two models of the 30S subunit and plot a matrix containing both error scaled and standard difference distance matrix elements put error scaled difference distance matrix elements to the lower left half of the displayed matrix 1 ddm seta 1 setb 2 esd_scaled on rbfind on 22 rbplot on ssplot off lolim 1 5 type lower limit 300 vx1 100 vy1 300 vx2 540 vy2 740 legend on legendx 100 legendy 220 legendw 180 ticks auto ticksfontsize 10 0 comments off pstype eps psfname ribo2_ddm eps title DDM native vs cognate par now put normal matrix into the other corner 1 ddm seta 1 setb 2 esd_scaled off rbfind off rbplot off ssplot off lolim 1 0 hilim 5 0 type upper legend on legendx 360 legendy 220 draw a diagonal for clarity ddm seta 1 setb 2 type diag I stop The matrix is shown in Fig 3 9 3 5 3 Superposition of molecules During the run of ribo1 inp ESCET generates two scripts that can be used superimpose the two models ribo1_o
6. 0 low and high limit for DD matrix psfname ddm_mers_all ps postscript output file stop Following is a script that will produce something similar to Figure 2 in the paper 6 i e all difference distance matrices in one plot The script is somewhat envolved I hope I will automate this in the future but I include it here in case you are facing a similar problem 16 Reading coordinates etc as in ddm_mers_all inp put first normal difference distance matrix ddm check loose ignore residue number for equivalences seta 1 setb 2 lolim 0 3 hilim 2 0 rb_find off vx1 100 vy1 560 vx2 170 vy2 630 bx1 50 by1 150 bx2 550 by2 750 Set the postcript Bounding Box type both esd_scaled off ticks auto comments off xticks off xtint 10 ticksfontsize 8 0 legend on legendx 100 legendy 200 legendw 180 pstype eps psfname ddm_mers eps Js and the other fourteen changing the viewport position as we go along We also need to worry about putting and not putting ticks ddm seta 1 setb 3 vy1 490 vy2 560 ticks auto xticks off legend off ddm seta 1 setb 4 vy1 420 vy2 490 ddm seta 1 setb 5 vy1 350 vy2 420 ddm seta 1 setb 6 vy1 280 vy2 350 xticks bottom ddm seta 2 setb 3 vx1 170 vy1 490 vx2 240 vy2 560 xticks off yticks off ddm seta 2 setb 4 vy1 420 vy2 490 ddm seta 2 setb 5 vy1 350 vy2 420 now the same procedure for the equivalent error scaled m
7. course you may need to put the correct coordinate transformation into the MOLSCRIPT input file first If you have the opengl version of MOLSCRIPT this is very easy With version 0 7 the use of LSQKAB for superposition has become obsolete The super imposed structures can be found in files called t_1DBF_ ABC _mr1 pdb A PYMOL script t_spos_mr1 pm1 is generated that displays the superimposed molecules In the first run lolim e 50 was used as a lower limit for significant changes To rerun the whole thing with a different setting of e 20 all we have to do is to change the lolim parameter of the ddm keyword and rerun the program The bit that is interesting in the log file can be found by searching the log file for the string conformationally invariant region Here the program tells us what it has identified as conformationally invariant gt The following residues form a conformationally invariant region A3 A13 A48 A52 A86 A95 There is some more useful stuff in that part of the file have a look around 3 2 Asparte Aminotransferase Here we read five models from five different files As the pdb files are a bit of a mess we have to put some of the numbers necessary for the calculation of uncertainties by hand Read the models ignore atoms with zero occupancy and generate coordinate uncertainties Soe Se ee ee 222 222 22222 aset_read ifile 7AAT pdb sele occ gt 0 0 aset_egen dmin 1 9 npar
8. ofile t_others pdb sele not protein and resn lt gt HOH stop 4 4 Comparing things 4 4 1 Compare two models The aset_comp command will try to do a very simple comparison of two atom sets WARN ING this is very a test aset_read tset 0 ifile coord1 pdb aset_read tset 1 ifile coord2 pdb sele element lt gt H aset_comp seta 0 setb 1 31 stop 4 4 2 Deriving some information from the sequence Sequences can be read from pir or pdb files For the latter SEQRES records are used if available otherwise the atom list is analysed Various numbers like the number of atoms will be printed seq_read ifile pdb 1HOE pdb 32 Chapter 5 Tips and Tricks 5 1 General 5 1 1 Why does LSQKAB not do what I want 1 Make sure that the FIT command is always before the MATCH command 2 Make sure that you are not using the same file for REFCRD and WORKCD 3 Make sure that there is no wrongly formatted metal atoms in the pdb file A typical error message for this is fmt read unexpected character apparent state internal 1 0 last format 6X I5 11X I4 lately reading sequential formatted internal 10 Abort 5 1 2 How do I use RASMOL There is a very good Quick REference Card at http info bio cmu edu Courses BiochemMols RasFrames REFCARD PDF Another website for RASMOL is at http www umass edu microbio rasmol An overview of a lot of RAS MOL related websites ca
9. rtype A sele resn ADE aset_amod set 2 rtype C sele resn CYT 4 1 3 Modifying a pdb file with double conformations to use it for MOLSCRIPT Making pictures of models that contain multiple conformations is not as simple as it could be MOLSCRIPT does not recognize alternate location identifiers in the pdb file so the usual trick is to split the pdb file into two files containing the different conformations and then to select atoms from the separate models The following ESCET script takes a pdb file applies an offset to all residuenumbers and then gets rid of all alternate location identifiers Please note that it is also necessary to sort the atoms as mixed up residue numbers confuse MOLSCRIPT aset_read ifile ref16r_1ls pdb read the original file aset_amod rnoffset 100 sele part B apply offset to all non B atoms aset_amod part sele part lt gt B remove all alternate location id s aset_sort set 0 aset_write ofile ref16r_1s_mod pdb stop 4 2 SHELXL related 4 2 1 Creating a file to test twinning operators at 3 A aset_read ifile xxx ins aset_write dmin 3 0 ofile xxx_mod ins otype res stop The resulting ins file will still need some editing but basically implements the following parametrization e residues as rigid bodies e two refinable B value per residue one for backbone one for sidechain e 3 81 A restraint on CA CA distance e SIMU 0 1 0 2 restraint for atoms c
10. the examples scripts Running the command escet h from the UNIX command line will create an HTML file called escet_ref html that can be used as a reference manual An ESCET script file can be executed from the UNIX command line using a command of the following form escet myscript inp gt myscript log amp Diagnostic output will then end up in myscript log graphical output will be dumped to a postscript file Scripts to look at the output using RASMOL and or MOLSCRIPT are put into files called myscript_out ras and myscript_out mol respectively Currently the program is only available as an executable for the Linux operating system If you find the program useful please cite the following paper Schneider TR A genetic algo rithm for the identification of conformationally invariant regions in protein molecules Acta Cryst D58 195 208 2002 I would also very much appreciate to receive reprints of papers that report the successful use of ESCET Thank you Please report all problems to me email thomas schneider ifom ieo campus it The program and its documentation are copyright Thomas R Schneider 2000 05 I would like thank the P testers in particular Karl Edman from Uppsala University for putting up with a semi functional version of the program and for making lots of useful suggestions Andrea Cocito IFOM Milan has been very helpful in tracking down some very nasty bugs Chapter 2 Command Scripts and Data Struct
11. 0 1200 1300 1400 1534 Figure 3 11 Plot of the distance between corresponding atoms in the 16S RNA after superpo sition based on the conformationally invariant part red and the uncertainty of the distance between atoms calculated by error propagation from the coordinate errors blue 27 Chapter 4 Utilities This chapter contains a loose agglomerate of ESCET scripts that are useful for all kinds of tasks All the functions described here are not really supported but may be useful anyway 4 1 Manipulating PDB files 4 1 1 Splitting a pdb file into separate chain id s ESCET script epi_split inp read coordinates of CA atoms from a pdb file prepare ten atom sets one for each molecule ic ea A A a e FAR Ia BR aset_select tset 1 sele chainid A aset_select tset 2 sele chainid B aset_select tset 3 sele chainid C aset_select tset 10 sele chainid J now write the all into different files aset_write sset 1 ofile t_A pdb aset_write sset 2 ofile t_B pdb aset_write sset 3 ofile t_C pdb aset_write sset 10 ofile t_J pdb stop 4 1 2 Change residue names If you need to change the residue names in a pdb file from 3 letter code to 1 letter code you can use the following commands in this case for atom set number 2 which happens to be a piece of 28 RNA aset_amod set 2 rtype G sele resn GUA aset_amod set 2 rtype U sele resn URI aset_amod set 2
12. 27968 aset_select tset 1 sele chainid A aset_read ifile 1TAR pdb sele occ gt 0 0 aset_egen nobs 36893 npar 26320 cp1 88 5 dmin 2 2 aset_select tset 2 sele chainid A aset_read ifile 1AMA pdb aset_egen dmin 2 3 npar 13892 nobs 17538 cp1 94 4 aset_select tset 3 sele all aset_read ifile 1TAS pdb sele occ gt 0 0 aset_egen npar 25408 nobs 17636 cp1 87 9 rfree 30 0 dmin 2 8 aset_select tset 4 sele chainid A aset_read ifile 1TAT pdb sele occ gt 0 0 aset_egen npar 25408 nobs 18194 cp1 97 0 dmin 3 0 rfree 30 0 aset_select tset 5 sele chainid A run difference distance matrix analysis ddm setl 1 5 esd_scaled on lolim 2 0 hilim 5 0 ssplot on rbplot on rb_iterate 3 stop The first run on all the models is used to decide which conformers are identical For this purpose a table is printed in the log file Para TATA lt esd gt A 011 032 0 25 0 29 0 39 JAAT A 100 0 733 76 2 819 ITAR A 819 84 5 86 7 1AMA 100 0 99 9 ITAS A 100 0 ITATA This table contains for every pair of models the percentage of elements in the error scaled difference distance matrix that are smaller than the threshold value of lolim 2 00 If for a pair 10 of models this value is larger than 98 0 the two models can be considered to be identical For an explanation see 6 This table can be used to decide which conformers are redundant and sh
13. 57 0 171 0 144 1 047 1IBL_A 1487 30S mRNA Cognate 52270 228284 89 9 27 5 3 11 dpiu 0 386 0 213 0 130 1 198 RNA Anticodon Stem loop Paramomycin Figure 3 6 Table in the HTML file summarizing the information pertinent to the estimation of coordinate uncertainties 21 Conformationally invariant part 1024 1486 atoms 68 9 A6 A31 A58 A76 A93 A285 A288 A303 A309 A352 A378 A382 A456 A474 A477 A481 A555 A610 A630 A910 A914 A940 A1025 A1035 A1039 A1043 A1062 A1195 A1237 A1260 A1271 A1298 A1340 A1351 A1368 A1405 A1409 A1492 A1498 A1516 A1519 A1533 Flexible part 462 1486 atoms 31 1 A32 A57 A77 A92 A286 A287 A304 A308 A353 A377 A383 A455 A475 A476 A482 A554 A611 A629 A911 A913 A941 A1024 A1036 A1038 A1044 A1061 A1196 A1236 A1261 A1270 A1299 A1339 A1352 A1367 A1406 A1408 A1493 A1497 A1517 A1518 Figure 3 7 Table in the HTML file showing the results of the search for the largest conforma tionally invariant region in 165 RNA A 1FJF X RAY DIFFRACTION 3 Ele Display Colours Options Settings Export a b Figure 3 8 Graphical display of the conformationally invariant region in blue based on the analysis following ribo1 inp a in RASMOL using the command rasmol script riboi_out ras b in MOLSCRIPT using the command molscript gl in ribol_out mol c in PYMOL using the command pymol t_spos_mr1 pml 3 5 2 Displaying difference distance matrices Normally
14. ESCET Version 0 7 User Manual Thomas R Schneider thomas schneider ifom ieo campus it February 28 2005 Contents 1 Introduction 2 Command Scripts and Data Structure 2 1 ATOM SOUS te Bose oe Ae ee ck eee SOS es ee ea eh E DD NOMS ax toe se ayes Sie dat Paley eat Ee A ie OO ee eh Ro en ee Se Sd 233 1 TRESIG WCS sia te O ee A ee aS NON 3 Identification of the rigid part of a macromolecule Sel Chorismat Mut se s a top ee 2 See eh ot es RS A aE g 3 2 Asparte Aminotransferase 2 2 2 ee 3 3 Tryptophan Synthase ee 3A Mersacidin u a ann OR a a a es 3 5 Ribosomal RNA 2 4 04 24 228mm a ann A 4 Utilities 4 1 Manipulating PDB files nn 4 2 SHEL XbL telated acid AA A 4 3 Analysing models ia anne Gite ESE nen nie Ree si 4 4 Comparing things 2 2 a 5 Tips and Tricks Bol General ss ar ed ran e a 52 ESGBT speche 4 222 ees hoi A ur ee a 6 Release Notes N a eR ww 10 14 16 20 28 28 29 30 31 33 33 34 36 Chapter 1 Introduction ESCET is a script driven program to analyse and compare three dimensional protein structures The central idea is to use estimated coordinate uncertainties throughout all calculations The underlying algorithms are described in a series of papers 6 7 8 This document contains a short introduction to the program The easiest way to get started is to look through the different scenarios discussed in chapter 3 and modify one of
15. aramters to be used in Cruickshank s equation 26 so at the moment number of parameters is set to 0 causing Cruickshank eq 27 the one based on Rfree to be used fixed mistake in URL given in program startup message fixed bug concerning y ticks Version 3 0 1c 1 Nov 2000 The manual was updated If the keyword check in the ddm command is set to auto the program will try to find a consistent set of atoms for the subsequent calculations all by itself This works in many cases but not always If it does not work the selection can still be done by hand using check loose or check strict I will try to make the algorithm more robust in the future so please send me examples were the automatic selection does not work Several atom sets can now be simultaneously selected by the set1 keyword of various commands The most important application of this is for difference distance matrices if you select a list of atom set via the setl keyword all pairwise difference distances will be printed Each matrix will be put onto a separate sheet producing a sort of book See example on page 16 Atoms can now be selected based on their element or number in their atom set This is very handy if you want to exclude hydrogens see example on page e g the following script will get rid of all hydrogen atoms found in a pdb file aset_read ifile test pdb aset_sele sele element lt gt H aset_write ofile test_without_hydrogens pdb A mu
16. are similar to each other with a mean rmsd for all 15 possible pairwise least squares superpositions using all C atoms of 0 83 A as calculated by LSQKAB 5 Analysis of the superimposed molecules to identify rigid and flexible regions was inconclusive For six molecules there are 6 x 6 6 2 15 possible pairwise comparisons The analysis of the corresponding 15 difference distance matrices allowed the indentification of rigid and flexible parts of the molecule The analysis is described in 6 For the six atom sets corresponding to the six molecules all possible pairwise difference distance matrices can be plotted on 15 pages of paper using the following script read coordinates and errors from SHELXL list file keep only CA atoms A A roi AEE aset_read ifile mers_fin_ls lst sele name CA prepare six atom sets one for each molecule A Se ee a aset_select tset 1 sele name CA and resi in 101 120 aset_select tset 2 sele name CA and resi in 201 220 aset_select tset 3 sele name CA and resi in 301 320 aset_select tset 4 sele name CA and resi in 401 420 aset_select tset 5 sele name CA and resi in 501 520 aset_select tset 6 sele name CA and resi in 601 620 automatically produce all error scaled difference distance matrices ddm setl 1 6 work on all six sets check loose loose consistency check esd_scaled on esd scaling is used lolim 3 0 hilim 5
17. arning is generated Version 8 0 6e 14 Jun 2004 e Section on the comparison of ribosomal RNA was added to the manual Version 8 0 6d 08 Jun 2004 e Notes can now be attached to models read from pdb files Use the notes keyword of the aset_read command to make notes e An overview of all conformers being analysed is now displayed as a table in the HTML output file Version 8 0 6 26 Feb 2004 e many small fixes Version 8 0 5 13 Oct 2003 e ESCET can now be started by giving the name of the inp as the first argument This will generate output files using filenames related to the input file name e In addition to the loggin information dumped to the Terminal the program now generates an output file in HTML format that can be printed cut and pasted etc e Several new entries have been added to the Utilities section of the manual Version 6 0 3 24 Jan 2003 e Loads of small fixes 37 Version p 0 2f 27 Jun 2003 e The Bounding Box for encapsulated postscript can now be set explicitly using bx1 by1 bx2 bx1 keywords of graphics related commands e Keywords xtdel1 and xtdel2 can be used to specify up to two tickmarks that will be explicitly deleted if e g if tickmarks overlap on the lower or upper end of a scale e Several problems with tickmarks for multi chain plots have been fixed e title and frametitle are now separate concepts frametitle is the title that will be put into the frame of a plot title is t
18. atrices ddm seta 1 setb 2 lolim 3 0 hilim 5 0 vx1 170 vy1 630 vx2 240 vy2 700 type both esd_scaled on limit 150 ticks off comments off xticks off yticks off legend on legendx 340 legendy 200 legendw 180 ddm seta 1 setb 3 vx1 240 vx2 310 legend off ddm seta 1 setb 4 vx1 310 vx2 380 ddm seta 1 setb 5 vx1 380 vx2 450 ddm seta 1 setb 6 vx1 450 vx2 520 ddm seta 2 setb 3 vx1 240 vy1 560 vx2 310 vy2 630 17 ddm seta 5 setb 6 vx1 450 vy1 350 vx2 520 vy2 420 ddm seta 1 setb 1 vx1 100 vy1 630 vx2 170 vy2 700 ticks auto yticks left ddm seta 2 setb 2 vx1 170 vy1 560 vx2 240 vy2 630 yticks off ddm seta 3 setb 3 vx1 240 vy1 490 vx2 310 vy2 560 ddm seta 4 setb 4 vx1 310 vy1 420 vx2 380 vy2 490 ddm seta 5 setb 5 vx1 380 vy1 350 vx2 450 vy2 420 ddm seta 6 setb 6 vx1 450 vy1 280 vx2 520 vy2 350 xticks bottom and we are finished Not so bad stop The figure produced Figure 3 5 represents a huge amount of information on a single A4 page 18 101 110 120 201 210 220 301 310 320 401 410 420 501 510 520 601 610 620 W A E 3 2 0A 0 3 0 3 2 0A 5 0 8 0 3005 00 Figure 3 5 Difference distance matrices and error scaled difference distance matrices for the six molecules of mersacidin In the lower left triangle ordinary difference distance matrices for all pairs of NCS copies are shown The color coding is according to the bar on the lower left all chan
19. bles escet script stats1 inp 30 calculate some statistics on a model 1 aset_read ifile feinschliff ref16s_ls lst sele element lt gt H and name lt gt W X3 aset_stat prop bfac label all full sele occ 1 0 label all partial sele occ lt gt 1 0 aset_select sele element lt gt H aset_stat prop bfac label all full sele occ 1 0 label all partial sele occ lt gt 1 0 label protein full sele protein and occ 1 0 label protein partial sele protein and occ lt gt 1 0 label water full sele resn HOH and occ 1 0 label water partial sele resn HOH and occ lt gt 1 0 label other full sele not protein and resn lt gt HOH and occ 1 0 label other partial sele not protein and resn lt gt HOH and occ lt gt 1 0 3 aset_stat prop esd label all full sele occ 1 0 label all partial sele occ lt gt 1 0 label protein full sele protein and occ 1 0 label protein partial sele protein and occ lt gt 1 0 label water full sele resn HOH and occ 1 0 label water partial sele resn HOH and occ lt gt 1 0 label other full sele not protein and resn lt gt HOH and occ 1 0 label other partial sele not protein and resn lt gt HOH and occ lt gt 1 0 we want to know what other really is so we write the atoms to a file aset_write
20. ch easier mechanism for selecting stretches of residues has been implemented e g selecting residues 19 to 25 36 to 42 45 and 56 to 58 can now be done in a statement of the form sele resi in 19 25 36 42 45 56 58 Some facilities to work with multiple models from NMR spectroscopy have been included The information about rmsd s from the NMR ensemble is translated into positional esd s using the aset_egen command Two models are available esd_model rmsd simply uses the rmsd as the coordinate uncertainty and esd_model rmsd2 uses the square of the rmsd as the coordinate uncer tainty In both cases the uncertainty is multiplied by the number given via the keyword esd_fac to allow to put the coordinate error on some pseudo absolute scale This facility is useful if you want to compare NMR structures and X ray structures See under the aset_egen command in the reference manual A schematic representation of secondary structure elements can now be included into difference distance matrix plots Helices are shown as open sheets as filled boxes Secondary structure can be defined using the aset_amod command see reference manual and section 2 3 1 If HELIX and SHEET records are found in a pdb file the information is used The keyword ss_plot of the ddm command is used to trigger plotting of the secondary structure For example see Figure 3 4 A number of small bugs have been fixed Version 6 0 1 11 Jun 2000 This is the very first beta
21. d uncertainties has to added manually For more see 7 Tryptophan Synthase Section 3 3 Two models of Tryptophan Synthase are compared Both normal and error scaled difference distance matrices are displayed for comparison For more on this analysis see 6 Mersacidin Section 3 4 How to display 36 matrices for two different ways of comparing 6 NCS related molecules at once For a change estimated standard deviations are available from a SHELXL Ist file Ribosomal RNA Section 3 5 Two models of the 165 RNA from the 30S ribosomal subunit are compared After some necessary massaging of the original pdb files difference Distance matrices are displayed and r m s d plots are prepared 3 1 Chorismat Mutase Comparison of three NCS related molecules Here is the script to read the data divide the atoms into NCS related copies and run the analysis aset_read ifile 1DBF pdb Generate standard uncertainties based on the information found in the pdb file As the completeness of the data was not present in the pdb file the value has to be put in manually aset_egen sset 0 esd_model dpiu cpl 92 0 divide the structure into three independent sets of atoms SP pel a pl AAA A APA pS pl dee aset_sele tset 1 sele chainid A aset_sele tset 2 sele chainid B aset_sele tset 3 sele chainid C calculate an plot the error scaled difference distance matrices for all pairs of molecules setl 1 3 lis
22. dues in principle the Insertion Id should be interpreted to obtain sensible results This special nomenclature causes an enormous amount of complications as most programs do not handle Insertion Id s properly and the results obtained become a mess A pragmatic solution to circumvent all these complications is to simply ignore all atoms that have an Insertion Ids This is done by ESCET The only difference with respect to an analysis of protein structures is that for the comparison of RNA the phosporus atoms are used as representatives of the corresponding nucleotides This fact can be indicated by choosing rmarker P in the aset_read command when a coordinate file is read see script below The following script will read the coordinate files assign coordinate error estimates and find the largest conformationally invariant region of the 165 ribosomal RNA ESCET input file ribol inp compare two models of the 30S subunit read the first structure and tell ESCET that phophorus atoms should be used as the central atoms of residues and that only atoms with non zero occupancy should be read aset_read ifile pdbifjf ent rmarker P sele occ gt 0 0 notes 30S stitle 1FJF Jy filename central atom of a residue is P ignore atoms with occupancy lt 0 0 explanatory note about this model short name for the model generate coordinate uncertainties aset_egen esd_model dpiu n
23. e chainid A and part J determine the conformationally invariant part dam seta 1 setb 2 compare sets 1 and 2 ssplot off do not plot secondary structure as this is not defined for RNA lolim 1 5 use a lower limit of 1 5 sigma dd_plot on stop The plot see Fig 3 10 is stored as a POSTSCRIPT file under the name ribo_dist1 ps To relate the distance between corresponding atoms to the uncertainties of the coordinates of the atoms involved a plot containing both the previous plot and a combined estimate of the coordinate uncertainties can be produced ESCET input file ribo_dist2 inp plot distance between native and the lsq fitted liganded structure vs residue number and display an estimate of the error of that distance read two pdb files and declare P atoms as residue markers aset_read ifile t_1FJF_A_mr1 pdb rmarker P 25 3 5 3 0 25 2 0 0 5 0 0 6 100 200 300 400 500 600 700 800 900 1000 1100 1200 1300 1400 1534 Figure 3 10 Plot of the distance between corresponding atoms in the 165 RNA after superpo sition based on the conformationally invariant part aset_egen esd_model dpiu ni 51927 nobs 238205 cp1 94 0 rfree 25 2 dmin 3 05 esd_fac 1 0 aset_sele tset 1 sele chainid A and name P and part aset_read ifile t_1IBL_A_mr1 pdb rmarker P aset_egen esd_model dpiu ni 52270 nobs 228284 cp1 89 9 rfree 27 5 d
24. e performed on different conformers of the same molecule e To be able to write complete models after analysis and superposition for visualization purposes and for further analysis a slight change in the general approach used in ESCET had to be introduced Previously the atoms e g CA s had to be selected before entering the ddm command Now this selection is done within the ddm command using the sele keyword with a standard selection string The default for this selection is to select CA atoms Thus for most users nothing will change except that they have to remove the explicit selection of CA atoms in earlier parts of ESCET scripts 36 Users interested in phophorus atoms as representatives of RNA structure can use sele name P to switch to phosphorus atoms People interested in a analysing complex mixtures of atoms can go back to the old be haviour by stating sele all e Scripts for visualization of results using pymol are now generated t_spos_mrl pml will show all conformers superimposed using rigid region number 1 t_spos_mr2 pml will shown the super position on rigid region number 2 etc e Various pieces have been added to the html file e g a table containing statististics about coordinate uncertainties of all conformers a table containing rmsd values for all least squares superpositions done etc There is also a hyper linked table of contents at the top of the file e When atoms with zero occupancy are read a w
25. ed Once atoms have been read sets can be modified by applying selection criteria command aset_select keyword sele All commands related to sets of atoms have a syntax of the form aset_ and are described in the reference manual 2 2 Atoms Atoms are the elementary unit of the program An atom has a number of properties that can be used and modified Examples of such properties are x y z coordinates atom name chain identifier chemical element B factor etc pp 2 3 Residues In the framework of ESCET a residue is a set of atoms that share the same chain identifier and the same residue number To count a residue must have at least one atom named CA If properties of entire residues e g the average B factor of the sidechain atoms have to be stored the information is kept with the respective C atom 2 3 1 Secondary structure information A typical property of a residue is the type of secondary structure element it is located in This information is stored as a single letter in the atom property ssid for secondary structure iden tifier of the respective C atom The single letter code corresponds to the Kabsch amp Sander notation as for example used in PROCHECK The codes and their meanings are given in Ta ble 2 3 1 code meaning B residue in isolated beta bridge E extended strand participates in beta ladder G 3 helix 3 10 helix H 4 helix alpha helix
26. ges in distances smaller than 0 3 are shown as gray differences in distances between 0 3 and 2 0 are shown using a color gradient where red stands for expansion and blue for contraction light colors represent small changes dark colors large changes all differences larger than 2 0 A are shown as full blue and full red respectively The blocks in the upper right triangle show the error scaled difference distance matrices for all pairs of molecules Here all differences lower than 3 0 times o Af are mapped to gray Changes greater than 3 0 and smaller than 5 0 times o Af are colourcoded using a scheme analogous to the one used for ordinary difference distance matrices 19 3 5 Ribosomal RNA ESCET can not only be used to analyse protein structures but also to compare RNA molecules using the phophorus atoms as representatives of the nucleotides Here the analysis of the RNA compenent of two structures of the 30S ribosomal subunit one without 1FJF and one with a drug bound 1IBL is described All ESCET input files mentioned in this section can be found in the inputs directory of the distribution 3 5 1 Comparing two model The models with PDB codes 1FJF and 1IBL contain several stretches of residues whose residue numbers do not follow the standard numbering scheme of i e consecutive residues have con secutive numbers Instead there are a number of residue numbers such as 129 or 190 that are repeated several times for these resi
27. ght triangle right legend and error scaled differ ence distance lower left triangle left legend matrices between two conformers of the 168 RNA shown in one matrix 3 5 4 Analysis of the superposition 1FJF and the moved version of 11BL can now be compared in terms of distances between phosphorus atoms ESCET input file riboil inp compare two models of the 30S subunit read the first structure and tell ESCET that phophorus atoms should be used as the central atoms of residues and that only atoms with non zero occupancy should be read aset_read ifile pdbifjf ent filename rmarker P central atom of a residue is P sele occ gt 0 0 ignore atoms with occupancy lt 0 0 notes 30S explanatory note about this model stitle 1FJF short name for the model generate coordinate uncertainties I aset_egen 24 esd_model dpiu ni 51927 nobs 238205 cp1 94 0 rfree 25 2 dmin 3 05 define the first conformer of the 165 RNA aset_sele tset 1 sele chainid A and part read the second structure 1 aset_read ifile pdblibl ent rmarker P sele occ gt 0 notes 30S mRNA Cognate RNA Anticodon Stem loop Paramomycin stitle 1IBL I generate coordinate uncertainties aset_egen esd_model dpiu ni 52270 nobs 228284 cp1 89 9 rfree 27 5 dmin 3 11 33 define the second conformer of the 16S RNA aset_sele tset 2 sel
28. he title that will be put on top of the actual figure Use keywords title and frametitle to set the values If several plots are overlayed in one figure the titles will be overlayed as well Setting title should help in such cases For rp plot the defaults are title frametitle something invented by the program Version 0 2e 17 Apr 2002 e Two keywords patchlen and minfraglen are now available to adjust the polishing of a solution of the rigid body search See escet_ref html for details e Diagnostics for automatic consistency checks has been improved Version 6 0 2d 11 Apr 2002 e Some problems with automatic consistency checking of atom lists were fixed e The Pairwise comparison table now contains a line with the mean estimated error for the atoms that are actually compared This allows to easily pick out the best determined atom set from a set of redundant conformers e STDOUT is now flushed regularly to make monitoring easier Version 8 0 2c 1 Mar 2002 e A warning is now given if a pdb file contains B values that are smaller or equal 0 0 such B values would mess up the error estimation e The manual is again available in pdf format e The plotting if DD matrices for multi chain problems is still not perfect loads of book keeping to do but at least does not hang anymore e The program will now give an estimate on the memory it will use If this estimate is close to the physical RAM of your computer try
29. i 51927 nobs 238205 cpl 94 0 rfree 25 2 dmin 3 05 define the first conformer of the 16S RNA aset_sele tset 1 sele chainid A and part 20 read the second structure 1 aset_read ifile pdblibl ent rmarker P sele occ gt 0 notes 305 mRNA Cognate RNA Anticodon Stem loop Paramomycin stitle 1IBL generate coordinate uncertainties aset_egen esd_model dpiu ni 52270 nobs 228284 cp1 89 9 rfree 27 5 dmin 3 11 define the second conformer of the 16S RNA aset_sele tset 2 sele chainid A and part determine the conformationally invariant part dam seta 1 setb 2 compare sets 1 and 2 ssplot off do not plot secondary structure as this is not defined for RNA lolim 1 5 use a lower limit of 1 5 sigma dd_plot on I stop Please note that a lower significance limit of e 1 50 is used instead of the standard value of 2 00 This is done to take into account the fact that the coordinate uncertainties of phosphorus atoms are smaller than for carbon atoms for which the dpiu error model is calibrated The results of the run are dumped into an HTML file that can be displayed with any Internet browser see Fig 3 6 and Fig 3 7 Graphical representations of the results can be displayed with RASMOL and MOLSCRIPT Fig 3 8 Nobs tree esd Angstroem tow 0 en sor ua 1FJF_A 1486 30S 51927 238205 94 0 25 2 3 05 dpiu 0 3
30. loser than 4 0 A only standard amino acids will be written to output file e for technical reasons atoms will be resorted N CA C O CB 29 4 3 Analysing models 4 3 1 Measuring distances between groups of atoms The following script first calculates the centers of mass of the different protein chains and then prints the angle between the vectors connecting AB and CD this angle was used to describe a homotetramer with 222 symmetry aset_read ifile ihfb pdb aset_vect sell protein and chainid A sel2 protein and chainid B sel3 protein and chainid C sel4 protein and chainid D If you select only two groups of atoms the distance between their centers of mass will be calculated here the distance between the centers of mass of the imidizale of His282 and the atoms of a PEP molecule connected by a r electron system aset_read ifile ihfb pdb aset_vect seli chainid A and resi 282 and sidechain and name lt gt CB sel2 chainid A and resi 500 and name C1 or name C2 or name C3 or name 0C1 X 4 3 2 Making a Histogram for Q angles Something you always have to put into an atomic resolution paper aset_read ifile refi6r_1s lst sele protein and part lt gt B rp_plot prop omega xtrotate 60 0 vxp 90 vyp 90 pstype eps stop 4 3 3 Derive some statistics about a model Another script to fill some ta
31. mber of models 35 Chapter 6 Release Notes Known Bugs e For multiple chain analyses the rigid body display on the bottom of the DD matrices is sometimes messed up However the calculations and the RASMOL and MOLSCRIPT files are correct Version 0 7 28 Feb 2005 Version 0 7 introduces a number of major improvements e The Kabsch Algorithm for least squares superposition 4 has been implemented The pro gram will now automatically superimpose the conformers analysed by applying the Kabsch algorithm to the different conformationally invariant regions The tranformed coordinates are written to pdb files for later use The pdb files are named tXXXX_C_spos_mr pdb where XXXX is the pdb code of a conformers C its chain name is a number indicating the conformationally invariant region used for superposition e The genetic algorithm for the identification of conformationally invariant regions can now be run in an iterative fashion This allows to not only find the largest conformationally invariant region but also other rigid parts of a molecule For example src kinase can be divided into its four domains using this approach see 8 for a explanation The default number of iterations is 3 and can be changed using the rb_iterate keyword in the ddm command Setting rb_iterate 0 should go back to the pre version 0 7 behaviour of the ddm command Please note that for various technical reasons the iterative analysis can only b
32. min 3 11 esd_fac 1 0 Xi aset_sele tset 2 sele chainid A and name P and part plot the distance between corresponding atoms as a function of residue number I rp_plot rset 1 IFJF is the reference sset 2 1IBL is the working modell prop dist property to plot aver p consider only P atoms color red color of the curve ymin 0 ymax 3 5 min max for Y axis xtdel1 1500 do no print tick at 1500 to avoid overlap pstype eps write an eps file I plot the uncertainty of the interatomic distance as a function of residue number 1 rp_plot rset 1 sset 2 as before prop esddist esd dist sqrt esd pos1 2 esd pos2 2 aver p consider only P atoms color blue color of the curve ymin 0 ymax 3 5 min max for Y axis 26 stop The resulting plot Fig 3 11 shows that for a substantial part of the molecule the differences in atomic positions after superposition based on the conformationally invariant part are on the order of the coordinate uncertainty Particularly interesting are the regions around residues 400 500 where the observed differences are significantly larger than the noise and the region around residue 1000 where despite the large differences between the coordinates these differences are not necessarily significant as the coordinate uncertainties are rather large blue curve 0 6 100 200 300 400 500 600 700 800 900 1000 110
33. n be found at http www rasmol org 5 1 3 How do I avoid total confusion when colouring superimposed models in MOLSCRIPT A good strategy is to first extract only the pieces of the moved files that you really want read tmp pdbihvy ent copy molO require in chain A and in amino acids delete tmp read tmp pdbihw3_mr pdb copy moll require in molecule tmp in amino acids and in chain A delete tmp 33 For colouring first colour everything as flexible and then mark the rigid blocks with another colour set residuecolour molecule mol rgb 1 0 0 35 0 35 flexible set residuecolour from 61 to 94 rgb 0 5 0 5 1 0 rigid set residuecolour from 99 to 104 rgb 0 5 0 5 1 0 rigid set residuecolour from 4133 to 4147 rgb 0 5 0 5 1 0 rigid The selection for the rigid pieces will only adress existing residues as this is in fact part of the algorithm an atom has to be existing in all models in order to have a chance to be recognized as rigid 5 2 ESCET specific 5 2 1 How do I put a difference distance matrix into a POWERPOINT presen tation All difference distance matrices are plotted in POSTSCRIPT format This causes Problems when the plots are included into POWERPOINT presentations i e the plots are not visible on non POSTSCRIPT devices I do not know of a program that really converts well between POSTSCRIPT and for example wmf Windows Meta Format The solution is to write the matrix that you want in eps fo
34. ndole 3 glycerol phosphate IGP to indole and glyceraldehyde 3 phosphate a reaction and the subsequent condensation of indole with serine to form tryptophan reaction 3 The reactions take place at two active centers which are separated by a distance of more than 25 A but nevertheless are precisely synchronized 1 In a study aimed at the understanding of the interaction between the two active sites crystal structures of the enzyme in complex with the substrate analogue 5 fluoroindole propanol phosphate TRPS pdb entry 1A50 and in complex with both F IPP and L serine TRPSY Y where A A stands for the amino acrylate that is formed at the P site under the experimental conditions chosen pdb entry 1A5S were determined 9 The complete difference distance analysis is described in 6 Read the models ignore atoms with zero occupancy and generate coordinate uncertainties AAA AA aset_read ifile 7AAT pdb sele occ gt 0 0 aset_egen dmin 1 9 npar 27968 aset_select tset 1 sele chainid A aset_read ifile 1TAR pdb sele occ gt 0 0 aset_egen nobs 36893 npar 26320 cp1 88 5 dmin 2 2 aset_select tset 2 sele chainid A aset_read ifile 1AMA pdb aset_egen dmin 2 3 npar 13892 nobs 17538 cp1 94 4 aset_select tset 3 sele all aset_read ifile 1TAS pdb sele occ gt 0 0 aset_egen npar 25408 nobs 17636 cp1 87 9 rfree 30 0 dmin 2 8 aset_select tset 4 sele chainid A
35. ould not be included in the rigid body analysis in order to not overweight such conformers If for a group of conformers all matrix elements are larger than 98 0with the lowest average esd into the subsequent analysis Here for the groups of 7AAT 1TAR and LAMA 1TAS 1TAT the matrix elements indicate that the respective conformers are the same within experimental error To find the most precise representative of each group we can consult another table st ar Now DI Ba Be ee ne JAAT A 6992 60850 961 dpin 0 107 0 048 0 030 0 502 27968 ITARA 401 6580 36893 885 ma 220 dpin 0 326 0 048 0 222 0 550 I 26320 19 4 IAMA 402 3473 17538 944 na 230 dpm 0 252 0149 0041 1 460 III 13892 159 ITASA 40 6372 17636 879 300 280 dpm 0 295 0173 0 041 0970 ITATA 39 6352 18194 97 0 300 3 00 dpin 0 392 0257 0051 1 708 As 7AAT and 1AMA have the lowest mean coordinate uncertainties we should continue the analysis with those two conformers as the best representatives of their groups To run the analysis for those two models only all we have to do is to change the atoms sets in the ddm command in the above script ddm setl 1 3 lt was setl 1 5 esd_scaled on lolim 2 0 hilim 5 0 ssplot on rbplot on J The main result of the analysis is Conformationally invariant part 275 399 atoms 68 9 _ A5 A12 Ad2 A226 A232 A319 Flexible part 124 399 atoms 31 1 A13 A41 A227 A231 A320 A410 Now yo
36. r 2000 x 2000 atoms will take 2000 x 2000 x 4 Bytes float 16 MBytes of memory Comparing 6 such models will give 16 matrices i e roughly 256 MBytes of memory are necessary to run such a problem But hey memory is cheap 34 5 2 4 How can I run ESCET on homologous i e non identical molecules models Extending ESCET to work on homologous structures is one of the things I plan to include in the future Right now however you would need to allign the molecules externally and then compare the matching segments Let s say you have molecules A and B and you have matches like A10 A20 lt gt B15 B25 A25 A30 lt gt B42 B47 A50 A70 lt gt B48 B68 Then you could do selections like aset_sele tset 1 sele resi in 10 20 25 30 50 70 aset_sele tset 2 sele resi in 15 25 42 47 48 68 and run a normal ddm analysis using the keyword check loose to allow non identical residue names for atoms being compared ddm check loose seta 1 setb 2 Please note that with check 100se it is not possible to do an interative analysis of your ensemble to find domains In fact to make the script run you also have to specify rb_iterate 0 ddm check loose rb_iterate 0 seta 1 setb 2 5 2 5 How much memory will my ESCET job require The necessary memory in units of Bytes can be calculated via natom x natom x nmodel x nmodel 2 nmodel x 8Bytes where natom is the number of atoms and nmodel the nu
37. rmat and then read it into the program GIMP http www gimp org When loading the file you can choose the resolution I found that 300 dpi is usually good enough If the resolution is too low you will see strange interference pattern in the matrix Then save the image in tif format and read it into Xv From XV you can then save the image in gif format gif is very good at storing plots with lots of straight edges in a very compact form This detour is necessary because GIMP does not support gif anymore and XV is not very good at converting POSTSCRIPT to a pixel format With newish versions of GIMP and POWERPOINT PNG is probably the best intermediate format 5 2 2 Why does the program gets confused by atom names used for my co factor Sometimes atoms of co factors have rather strange names One examples is NADP The pdb says about this at the following location http www rcsb org pdb docs format pdbguide2 2 guide2 2_frame html in large het groups it sometimes is not possible to follow the convention of having the first two characters be the chemical symbol and still use atom names that are meaningful to users A example is nicotinamide adenine dinucleotide atom names begin with an A or N depending on which portion of the molecule they appear in e g AC6 or NC6 AN1 or NN1 5 2 3 After I start ESCET my computer starts swapping Why As it stands the program is not really optimized in terms of memory consumption A matrix fo
38. rtunate situation of having a SHELXL Ist file from a full matrix inversion available some estimates for the coordinate error have to be generated Normally one would use the B factor scaled DPI explained in 6 aset_egen command The program is pretty clever in extracting the information it needs to calculate the error estimates from the pdb file In case not all the necessary pieces can be found the information can be added by hand see section 3 2 Define the conformers that should be compared In complicated cases one can also restrict selections to chainids and residue ranges etc to obtain consistent sets of atoms the program tries to figure out the largest consistent set itself but sometimes needs a bit of help aset_select command Run the automated error scaled difference distance matrix analysis ddm command Look at the results in the html file or use the automatically generated scripts for LSQKAB MOLSCRIPT RASMOL and PYMOL to create some more intuitive represen tations The following examples are discussed Chorismat Mutase Section 3 1 Compare three NCS related copies of Chorismat Mutase at 1 3 Make figures that can be used for publication in Acta Cryst D For more see 7 Aspartate Aminotransferase Section 3 2 Compare five different models of Aspartate Aminotransferase The models have been refined to different resolutions and some of the information necessary to generate standar
39. t of atom sets to look at esd_scaled on use error scaling lolim 5 0 hilim 10 0 limit for matrix xtint 50 ticks suitable for publication ticksfontsize 14 font suitable for Acta after shrinking ssplot on plot secondary structure rb_plot on plot rigid body description pstype ps produce an ps file on output psfname cmut_ddm_5 0 ps filename for graphics finito stop To run this script type gt escet cmut_ddm_5 0 inp gt cmut_ddm_5 0 log This will dump the difference distance matrices into a file called cmut_ddm_5 0 ps You can look at this file with GHOSTVIEW if your are interested Otherwise you can check out the log file cmut_ddm_5 0_out html Most interesting is probably the graphical representation of the results files called cmut_ddm_5 0_out ras and cmut_ddm_5 0_out mol have been created that can be used to show the results using RASMOL and MOLSCRIPT respectively Typing rasmol script t_ras ras will bring up a nice little RASMOL window displaying your molecule with the conformationally invariant atoms in blue and the flexible ones in red see Figure 3 1 a a b Figure 3 1 a RASMOL display of a molecule of Chorismate Mutase with rigid parts mapped onto the molecule in blue flexible parts in red b The same with MOLSCRIPT Typing molscript r lt t_molscript mol render jpeg gt t_molscript jpg will produce a rendered molscript plot See Figure 3 1 b Of
40. test release of ESCET Please be patient and report all errors problems glitches misunderstanding etc p p to trs at shelx uni ac gwdg de 40 Bibliography 1 K S Anderson E W Miles and K A Johnson Serine Modulates Substrate Channeling in Tryptophan Synthase J Biol Chem 266 8020 8033 1991 2 D W J Cruickshank Remarks about protein structure precision Acta Cryst D55 583 601 1999 3 C C Hyde and E W Miles The Tryptophan Synthase Multienzym Complex Explor ing the Structure Function Relationships with X ray Crystallography and Mutagenesis Bio Technology 8 27 32 1990 4 W Kabsch A solution for the best rotation to relate two sets of vectors Acta Cryst A32 922 923 1976 5 W Kabsch A discussion of the solution for the best rotation to relate two sets of vectors Acta Cryst A34 827 828 1978 6 T R Schneider Objective comparison of protein structures error scaled difference distance matrices Acta Cryst D56 714 721 2000 7 T R Schneider A genetic algorithm for the identification of conformationally invariant regions in protein molecules Acta Cryst D58 195 208 2002 8 T R Schneider Domain identification by iterative analysis of error scaled difference distance matrices Acta Cryst D60 2269 2275 2004 9 T R Schneider E Gerhardt M Lee P Lian K S Anderson and Schlichting I Loop closure and intersubunit communication in tryptophan synthase Biochem
41. to use a mashine with more memory to avoid swapping 38 Version 0 2 1 Feb 2002 e Rewrote documentation to reflect zillions of changes e Major improvements on RASMOL MOLSCRIPT and LSQKAB scripts generation e Substantial cleanup on the log file Version 8 0 1g 5 Jun 2001 e Fixed some problems with MOLSCRIPT output of the rigid body finder e Argument dd_plot off in keyword ddm will switch off the postscript output Version 0 1f 25 May 2001 e The conformationally invariant part of a molecule can now be automatically determined using a ge netic algorithm Right now no parameters are accessible from the interface changing parameters still has to be done in the source But using rb_find on in the ddm command will run the algorithm using a reasonable set of default parameters Setting rb_plot on and ss_plot on will put the information also into the plotted difference distance matrices rigid parts are marked as dark gray flexible as light gray The program also dumps a bunch of files to use as input to MOLSCRIPT RASMOL and LSQKAB I will work hard on documenting these promised molscript gl lt t_molscript mol will give a pretty picture in many cases blue for rigid red for flexible If it doesn t work it will at least be a good template If more than two atom sets were selected using the setl keyword in the ddm command the auto matic interpretation will take all models into account automatically
42. u can use the corresponding Cy atoms in your favourite fitting program Alternatively ESCET has written a number of PDB files named t_1TAT_A_mr1 pdb etc whereby mr1 stands for moved using rigid region number 1 which contain all conformers superimposed onto the reference conformer and a PYMOL script that will display the superimposed molecules Running pymol t_spos_mr1 pml 11 PyMOL gt _ Mouse enu Menu Cent PkAt M P 1 O sec eee Figure 3 2 PYMOL displaying the result of the analysis of the analysis of the difference distance matrices as a set of superimposed conformers The first conformationally invariant region is shown in blue and used for superposition the second in green Flexible parts of the molecule are shown in red will create the following view shown in Fig 3 2 As for the previous examples there are some premade RASMOL and MOLSCRIPT scripts Running molscript r in aatase_ddm2_out mol render jpeg gt aatase_ddm2 jpg will give you a plot like the one in Figure 3 3 12 Figure 3 3 MOLSCRIPT figure of a molecule of Aspartate Aminotransferase with rigid parts mapped onto the molecule in blue flexible parts in red For this figure the script t_molscript mol generated by ESCET the only editing done concerned the orientation matrix 13 3 3 Tryptophan Synthase Tryptophan synthase catalyses the last two reactions in the biosynthesis of tryptophan the cleav age of i
43. ure The program stores data pertaining to atoms in atom sets An atom set can contain an almost unlimited number of atoms the number of sets is currently limited to 250 Atom sets can be manipulated and analysed using a large number of commands The commands are normally put in a command script which is then run from a UNIX shell A typical command script has the form ESCET script for making a Bfactor plot read information about atoms from pdb file 21zt pdb into atom set number one aset_read tset 0 ifile 21zt pdb select atoms in residues 5 to 100 Use atom set number 0 as the source and atom set number 1 as the target aset_sele sset 0 tset 1 sele resi in 5 100 make a plot of the property bfac of the CA atoms vs residue number for the atoms found in sset 1 rp_plot sset 1 prop bfac aver ca Some output will be printed to the screen graphical output will be dumped to a POSTSCRIPT file 2 1 Atom sets Atom sets represent the information that can be found for example in a pdb file They are basically a collection of atoms with some general information such as name of the protein or the crystallographic unit cell found in the pdb file attached Information about atoms can be read from from various file formats Currently supported formats are pdb SHELXL ins res SHELXL lst As much information as possible is extracted from the respective files e g if an lst file contains s u s these are stor
44. ut_lsqkab1 csh is the script that will use LSQKAB to superimpose models using the atoms found to be conformationally invariant in the analysis WARNING The version of LSQKAB coming with CCP4 4 2 can only handle 20000 atoms Thus the 30S structure can not be handled There are two solutions 1 work only with a subset of the coordinate file such as only P atoms or 2 recompile LSQKAB after changing the parameter NATOM in the header of 1sgkab f WARNING RASMOL will crash when reading large coordinate files written by LSQKAB The reason for this is that the CONECT records in the PDB file are messed up The simplest solution is to delete the CONECT records from the file The second command file generated by ESCET ribo1_out_lsqkab2 csh will do the superpo sition based on atoms that are part of the conformationally invariant region and which have a coordinate uncertainty lower than a threshold in the reference model The default of the thresh old is 0 2 A the value can be changed using the RigLoLim keyword of the ddm command To make this script usable follow the instructions in its header Running ribo1_out_lsqkabi csh will superimpose 1048 phosphorus atoms with and r m s d of 0 67 A The moved model is in a file called pdb1ib1_p_mr pdb 23 a S S lhi ai el ES E ao ane ToT al 6 200 300 400 500 600 700 800 900 1000 1100 1200 1300 1400 1588 3 80 150 150 3 80 5 0A 1 0 1 0A 5 0A Figure 3 9 Elements of the standard upper ri
Download Pdf Manuals
Related Search