The EMfit program

Setting up the Control File (1) (input and output files) INPUT Line 1. Path to cryoEM map. Line 2. Path to X-ray map to be used as a scaling standard. Line 3 Path to structure to establish mask within which maps should be comparable. OUTPUT Line 4 Path to output maps: map1, map2, or (map1 - map2).

/bio/indiana/mgr/EMmaps/corr_symm_219_368_sc.em /u19/mr/RRvfit/E1.1/mono.pdb /u19/mr/RRvfit/E1.3a/mono_restrn_tight.pdb /u19/mr/RRvfit/E1.3a/mono_target+318.pdb /pucc/scratch6/mr/junk-1.pdb /pucc/scratch6/mr/junk-2.pdb /pucc/scratch6/mr/junk-3.pdb /pucc/scratch6/mr/junk-4.pdb /pucc/scratch6/mr/junk-5.pdb end ~

Setting up the Control File (2) (map properties) Line 5 (mapform(i),i=1,4) Map formats (no input, Purdue EM, Purdue crystallographic, CCP4, convenient ASCII) Use T.Dokland Maul program for interconvertion, available on Purdue ftp server and on web. Line 6. (scale(i), i=1,3) Scale factors to multiply all densities on each of the three input maps. Note : EM densities conventionally show molecule with negative density

/bio/indiana/mgr/EMmaps/corr_symm_219_368_sc.em /u19/mr/RRvfit/E1.1/mono.pdb /u19/mr/RRvfit/E1.3a/mono_restrn_tight.pdb /u19/mr/RRvfit/E1.3a/mono_target+318.pdb /pucc/scratch6/mr/junk-1.pdb /pucc/scratch6/mr/junk-2.pdb /pucc/scratch6/mr/junk-3.pdb /pucc/scratch6/mr/junk-4.pdb /pucc/scratch6/mr/junk-5.pdb end ~

Setting up the Control File (3) (input map properties) Lines 7, 8 & 9 (orig1(i), i=1,3), a1,hand(1),da1, na1,jturn(1) (orig2(i), i=1,3), a2,hand(2),,jturn(2) (orig3(i), i=1,3), a3,hand(3),,jturn(3),ismear,jsmear The pixel position of the origin (e.g.virus center). Can be fractional (e.g. 20.5), representing a point between pixels a, da and na is size of pixel used for for. scaling hand (+ or – 1); jturn (0 no, 1 turn 90deg.) ismear, jsmear used for mask determination.

/bio/indiana/mgr/EMmaps/corr_symm_219_368_sc.em /u19/mr/RRvfit/E1.1/mono.pdb /u19/mr/RRvfit/E1.3a/mono_restrn_tight.pdb /u19/mr/RRvfit/E1.3a/mono_target+318.pdb /pucc/scratch6/mr/junk-1.pdb /pucc/scratch6/mr/junk-2.pdb /pucc/scratch6/mr/junk-3.pdb /pucc/scratch6/mr/junk-4.pdb /pucc/scratch6/mr/junk-5.pdb end ~

Setting up the Control File (4) (map scaling) Line 10 radout, radin, rinc. nvar, rlimit, radrna, crit1, iverbose, local Determine scale for density between radout and radin where it is consistent with density of mask defined by (X-Ray) map 3. Use crit1 to define mask. Determine CC and R factor for sequential pixel sizes to find correct magnification r2 = a + b*r1: one or two param scale factor

/bio/indiana/mgr/EMmaps/corr_symm_219_368_sc.em /u19/mr/RRvfit/E1.1/mono.pdb /u19/mr/RRvfit/E1.3a/mono_restrn_tight.pdb /u19/mr/RRvfit/E1.3a/mono_target+318.pdb /pucc/scratch6/mr/junk-1.pdb /pucc/scratch6/mr/junk-2.pdb /pucc/scratch6/mr/junk-3.pdb /pucc/scratch6/mr/junk-4.pdb /pucc/scratch6/mr/junk-5.pdb end ~

Setting up the Control File (5) (map output) Line 11 Maptype, mapstart, mapstop Initial and final sections output as map or difference map Log file contains simple representation of map See example for polio virus in file pub/mgr/Polio/polio.out on Purdue Server

/bio/indiana/mgr/EMmaps/corr_symm_219_368_sc.em /u19/mr/RRvfit/E1.1/mono.pdb /u19/mr/RRvfit/E1.3a/mono_restrn_tight.pdb /u19/mr/RRvfit/E1.3a/mono_target+318.pdb /pucc/scratch6/mr/junk-1.pdb /pucc/scratch6/mr/junk-2.pdb /pucc/scratch6/mr/junk-3.pdb /pucc/scratch6/mr/junk-4.pdb /pucc/scratch6/mr/junk-5.pdb end ~

Manipulation of Electron Microscopy Density MAP INPUT AND OUTPUT Note: The maps are assumed to be sampled on an orthogonal grid referred to the "map" coordinate system. Line 1. path to first input map file (unit 1) Line 2. path to second input map file (unit 2) Line 3 path to third input map (file (unit 7) (determines mask) Line 4. path to output difference map (map1 - map2) Line 5. (mapform(i),i=1,4) (free) mapform(i)=0 no input for this map =1 Purdue EM format =2 Purdue crystallographic write3d format =3 CCP4 map format =4 convenient ascii for output on unit 6 =5 Purdue EM format output for a difference map i=1,4 corresponds to the maps given on lines 1 to 4. Line 5b. If mapform(4)=5 then read molone, moltwo (free) This will output onto output tape 4 the difference density between molecule_1 and molecule_2. It considers every grid point within a radius of rdif (line ) of all atoms in molecule_1 with respect to the corresponding (interpolated) densities in molecule_2 as given by the NCS matrecies. Output will be on tape 4 as Purdue ASCII EM format. Line 6: (scale(i),i=1,3),scale(4) (free) scales to be applied to each of the three input maps. These mutliply the densities at each grid point. Note that the E/M density convention is to let the molecule have negative density. Hence it is usually useful to make this scale factor negative. (default is -1.0) scale(4).eq.0 if output pdb file is to contain temperature factors based on interpolation of density at fitted atom site : (100.0-rho)/bigmost.ne.0 if output pdb file is to contain temperature factors based on sum of density within sphere of radius RDIF at each atom : (sumrho*100.0)/sumrhomax The temp factor is set to zero when sumrho<0.0.

Line 6: (scale(i),i=1,3),scale(4) (free) scales to be applied to each of the three input maps. These mutliply the densities at each grid point. Note that the E/M density convention is to let the molecule have negative density. Hence it is usually useful to make this scale factor negative. (default is -1.0) scale(4).eq.0 if output pdb file is to contain temperature factors based on interpolation of density at fitted atom site : (100.0-rho)/bigmost.ne.0 if output pdb file is to contain temperature factors based on sum of density within sphere of radius RDIF at each atom : (sumrho*100.0)/sumrhomax The temp factor is set to zero when sumrho<0.0.

Line 7: (ORIG1(i),i=1,3),A1,HAND(1),DA1,NA1,JTURN(1),MVORIG(1) (free) orig1(1),orig1(2),orig1(3) is the pixel position of the origin in the first map (the virus center). It should be given as a real number as sometimes the origin is not on but between pixels. The origin position is counted in pixels with the first pixel being at position 1, a1 is the distance between pixels in the first map hand(1) designates the hand of the first map. If if it is.lt.0 the map is reflected through the plane perpendicular to y passing through orig1 after reading the map. NOTE: The map is inverted on first reading it in. Thus to use the real hand set hand=-1, except when using jturn=2 (see below) da1 is the increment in a1 to be tried for na1 steps to test for the best value of a1 relative to a2. that is all values of a1 to a1+da1*(na1-1) will be tried in steps of da1. If na1=0 then only the value of a1 will be tried. jturn(1) turn map through 90 deg. =0 do not turn =1 Turn by 90 degrees about the position designated as origin (orig1) ii=orig1(1)-(j-orig1(2)) and jj=orig1(2)+(i-orig1(1)) ii=orig1(1)+orig1(2)-j and jj=orig1(2)-orig1(1)+i =2 turn through 90 degrees about the pixel at (0,0) of each section, AND change the hand. Note that the hand is altered when the map is first read. (ii=j and jj=i) =3 turn through 90 degrees about pixel at (0,0) of each section. Do NOT change hands. (ii=j and jj=ncol-j) mvorig(1) This is primarily to determone where the middle of a print of the section will be. =0 or 1 set it at (nsec+1)/2 =2 set it orig =3 set it at orig+1 =4 if the map contains a whole virus then read only the second half of the map (sections go from (say) -150 to In essence all sections below ORIG(3) are read but not stored. Storing starts with section containing the ORIG(3) point.

Line 8: (ORIG2(i),i=1,3),A2,HAND(2),JTURN(2),MVORIG(2) (free) orig2(1),orig2(2),orig2(3) is the pixel position of the origin in the second map (the virus center). It should be given as a real number as sometimes the origin is not on but between pixels. The origin position is counted in pixels with the first pixel being at position 1, a2 is the distance between pixels in the first map hand(2) is +/-1.0 to designate the hand of the second map. If it is -1.0, the map is reflected through the plane in z through the center of the virus. jturn(2) turn map through 90 deg. mvorig(2) move origin from a corner to center of the map (see line 7) Line 9: (ORIG3(i),i=1,3),A3,HAND(3),JTURN(3),ISMEAR,JSMEAR,MVORIG(3) (free) Same as Line 8 but for third map. ismear is distance (I+/-ismear) in 3D used for smearing the density of map3 to define the mask. jsmear =1 for RMS dev from mean =2 for max density mvorig(3) move origin from a corner to center of the map (see line 7)

MAP SCVALING OPERATIONS Line 10: RADOUT,RADIN,RINC,NVAR,RLIMIT,RADRNA,CRIT1,IVERBOSE,LOCAL radout radin. These are the outside,inside limits for determening the horizontal and vertical scales. rinc Used to analyse effect of applying horizontal scaling to vertical scaling in annuli. Annuli will have inside radii 0.0,rinc,2*rinc,.. outside radii rinc,2*rinc,3*rinc,.. till the outside radius is bigger than rlimit. default=50.0 CRIT1.eq.0 horizontal scaling computed from denseties between radout and radin. vertical scaling in radial shells according to rinc between radout and radin, extrapolated to rlimit. If rinc.eq.0 then do horizontal scaling only but show vertical scaling although use results only from horizontal scaling. CRIT1.gt.0 do horizontal scaling on those denseties for which (rho2-ave2)>(crit1*rms_rho2). Show vertscaling in shells only for stat info. Also used in out3d to define protein mask when (rho(2)-ave(2))>(crit1*rms(2)) and to define icam density when (rho(1)-ave(1))>(crit1*rms(1)) nvar (=1 or 2 with default=1) the number of variables to be used for scaling. The scale factor to be applied onto the 2nd map (Xray map) when n=1 scale factor is a+b*rho2 2 b*rho2 rlimit Set all density to zero beyond this radius, to avoid noise from one or the other map in the external solvent region(default=1000.0) radrna Now used only in out3d to draw a sphere (+) on printer output iverbose verbosity of output: 0 least verbose 1 intermediate 2 intermediate 3 most verbose Note: If iverbose.ge.2 then there will be single character output for map or difference map on standard output, unit 6. But note footnote at end of details. local apply local vertical scaling (0=no, 1=yes) horizontl scaling in fact does both hor. & vert. scaling using density between radin and radout. Then vert. scaling uses the best horizontal scaling in annuli. If local.ne.0 then vert scaling is applied in radial shells.

Line 11a: MAPTYPE,MAPSTART,MAPSTOP,ICOS,IWRSTR,KDOM,INHEAD (free) maptype=0 or 3 print difference map (map1-map2) =1 or 4 print out map 1 in routine out3d =2 or 5 print out map 2 in routine out3d when maptype=3,4,or 5 the print out will place an x in regions outside the mask. mapstart, mapstop the first and last sections to be printed, setting the origin section as section zero. icos if icos.ne.0 it at the limits of the icos triangle assuming that the 2fold to 5-fold direction is along y. NOTE: to save space the pgm uses only the negative sections the zero section up to section "isec" set as a parameter. iwrstr can be used to aline symmetry axes uusually set iwrstr=0. Otherwise: =1 rotate input coords 180deg about x..eq.1 or 3 limit search angles by aligning symmetry axes parrallel.eq.2 or 4 limit search angles by aligning symmetry axes anti-pllel When iswrstr= 3,4,5, or 6 rotate molecular diad to coincide with the u2,v2,w2 direction. (see line 11b for special input) kdom=0 Assume all atoms are in the same domain kdom<0 read in a list of domain limits and special markers for labelling atoms or atom regions on maplot output for each retained fit. Input consists of (limits(nter,i),i=1,4) (free) where limits(inter,1)>=domain number (1,2,3,...) <=character to be used on maplot -1 for A, -2 for B, etc limits(inter,2)= first residue number in domain limits(inter,3)= last residue number in domain limits(inter,4)= integer weight to be applied to residues in this domain note that if there are more than one set of atoms in a domain, both sets can be entered on seperate lines but with the same domain number The same residue(s) can be mentioned in a domain description and for marking on maplot. inhead =0 then format for the second line of the header is 3i4,5f8.3 =1 then format for the second line of the header is 3i4,1x,5f8.3

Line 11b: if((iwrstr.eq.3).or.(iwrstr.eq.4)) thean read: U1,V1,W1,U2,V2,W2 (free0 where U1,V1,W1 are the direction cosines of the direction in the atomic model (after pre-rotation) that is to be moved onto U2,V2,W2 in the map. Thus the atomic co-ordinates when read in are: 1. pre-rotated usually so as to put a special 2-fold axis parallel to the z-axis (u1,v1,w1) by multiplying with [prerot]. 2. These coordinates are then rotated by the Eulerian angles theta1, theta2, theta3 by multiplying with [arot]. 3. Next the u1,v1,w1 axis is placed onto axis u2,v2,w2 by rotation about u3,v3,w3, by muttiplying with [relign]. 4. finally the present position of the molecule is repeated by means of the NCS operators [rotp] to give the final matrecies [rothp].

Setting up the Control File (6) (real space atom fitting) Line 12 icont, isearch, iclimb, iturn, ntop, ifit icont= 0 (no search), 1 (Ca only), 2 (all atoms) isearch=0/1 complete Eulerian search iclimb=0/1 “climb” on top results (to be shortly augmented by a least squares procedure) ifit: 3 choices for best fit criterion See example for Ross River Virus in file pub/mgr/Sindbis/rrv.out

/bio/indiana/mgr/EMmaps/corr_symm_219_368_sc.em /u19/mr/RRvfit/E1.1/mono.pdb /u19/mr/RRvfit/E1.3a/mono_restrn_tight.pdb /u19/mr/RRvfit/E1.3a/mono_target+318.pdb /pucc/scratch6/mr/junk-1.pdb /pucc/scratch6/mr/junk-2.pdb /pucc/scratch6/mr/junk-3.pdb /pucc/scratch6/mr/junk-4.pdb /pucc/scratch6/mr/junk-5.pdb end ~

Line 12: ICONT,ISEARCH,ICLIMB,ITURN,NTOP,IFIT icont determines nature of refinment for fitting atoms into the map =0 then this is the end of the job. Dont fit anything. =1 use Calpha atoms only =2 use all atoms in order to maximize the electron density height at all atoms. =5 stop program after determening height of density at target sites atom sampling: if icont<10 interpolate at atom position if icont=icont+10 take mean of all grid points within a radius of rdif for each atom isearch=0/1. Do not or do a systematic, point by point search The pgm will retain the best fits and feed it into the climb routine if iclimb=1. =1, 4 or 6 search complete asymetric unit of Eulerian space =2, 5 or 7 search only within limits set by input =3 explore the best orientation of a NCS symmetry element given the stheta1,stheta2,stheta3 angles This requires input on Line 21. Note that this function only works with Ncryst operators, not with NCS operators. Thus ncs must be set to 0 on line 17. if isearch is 4,5,6, or 7 use search6d. For each angle combination do a 3d xyz climb if isearch is 4 or 5 use only fsum and neg (-den) to compute rcrit in search6.f for climb. (This is fast) if isearch is 6 or 7 use fsum and neg (-den) nclash and rmsdist to compute rcrit in search6.f for climb. (This is slow but better) iclimb =0/1. Do not or do a systematic, point by point climb = 2 after completion of the climb (isearch=0,iclimb=2), or search and climb (isearch=1 or 2, iclimb=2), then search in a volume of +/-2 increments around the best fit to determine the accuracy and sensitivity of the fit. Note: the criterion of fit is the mean density at all atoms * 100.0/biggest, where biggest is the largest density in the map

iturn =0/1,2Do not or do rotate output coordinates by 90deg about z to make the EM and Xray system match. 1. x'= y (turn about defined origin) y'=-x z'= z 2. x'= y (turn about (0,0,0) of map and invert) y'= x z'= z ntop Climb on the ntop 'peaks' from the prior search results if isearch>0. Otherwise climb on the site given by the input in lines 11 and 12. criterion for quality of fit over all sampled atoms: ifit =0 or 10 use mean density of atoms as search criterion. Search for a maximum =1 or 11 use rms density for rho>0 as search criterion Search for a maximum =2 or 12 use rms scatter of fitted density. Search for a minimum =3 or 13 use domain weighted mean density of atoms as search crit Search for a maximum if ifit.lt.10 normalize output sumf by the biggest map value ge.10 normalize by the rms deviation from the mean of the map

Setting up the Control File (7) (3D Eulerian search) Line 13 stheta1, stheta2, stheta3, dtheta1, dtheta2, dtheta3, ftheta1, ftheta2, ftheta3, Start (s), increment (d), finish (f) of Eulerian search Limits are set for complete search if icont=1 Above limits are used if icont=2 ntop results are retained for further refinement iverbose determines whether complete search, ntop results only, or nothing is copied to log.

/bio/indiana/mgr/EMmaps/corr_symm_219_368_sc.em /u19/mr/RRvfit/E1.1/mono.pdb /u19/mr/RRvfit/E1.3a/mono_restrn_tight.pdb /u19/mr/RRvfit/E1.3a/mono_target+318.pdb /pucc/scratch6/mr/junk-1.pdb /pucc/scratch6/mr/junk-2.pdb /pucc/scratch6/mr/junk-3.pdb /pucc/scratch6/mr/junk-4.pdb /pucc/scratch6/mr/junk-5.pdb end ~

Matrix Algebra Let [E] be the rotation matrix that takes the atomic coordinates from the PDB axes to a specific orienation and site in the EM density. Thus x’ = [E]x + d [E] must be explored or refined and is defined in terms of three angles, usually Eulerian angles. Let [R n ] define the nth position of the atomic structure in the target EM density. Thus x’’ = [R n ]x’, and x’’ = [R n ][E]x + d n. [R n ] are the known NCS operators.

Setting up the Control File (8) (3D climb refinement) Line 14 Centerx, centery, centerz, dx,dy,dz, fxyz1, fxyz2, fxyz3 Initially the C. of G. of the atom list will be placed on centerx, centery, centerz. The maximum allowed translation limits are centerx +/- fxyz1, etc. After the first climb finds improved angles and positions, all increments are devided by 10 and climb is repeated. And so forth until all increments are <0.25deg and <0.5A Climb is performed on all ntop search results.

/bio/indiana/mgr/EMmaps/corr_symm_219_368_sc.em /u19/mr/RRvfit/E1.1/mono.pdb /u19/mr/RRvfit/E1.3a/mono_restrn_tight.pdb /u19/mr/RRvfit/E1.3a/mono_target+318.pdb /pucc/scratch6/mr/junk-1.pdb /pucc/scratch6/mr/junk-2.pdb /pucc/scratch6/mr/junk-3.pdb /pucc/scratch6/mr/junk-4.pdb /pucc/scratch6/mr/junk-5.pdb end ~

Climb Refinement rcrit = nclash * (-den), is the basis of rank rcrit is normalized with respect to the top fit prior to climb refinement nclash are the number of atoms that are less than rdif Å from any other atom in a neighboring subunit -den are the number of atoms in negative density sum is the sum of densities for all atoms in one NCS assymetric unit.

Setting up the Control File (9) (Atomic co-ordinates) Line 15 (INPUT) Path to model coordinates used for fitting Lines 16 (OUTPUT) Paths to files for keeping the n top fits. End the list of files with a final line containing the word ‘end’.

Restraints Line 15a path to atomic model coordinates input file (PDB) Line 15b path to PDB file of restraining atoms The temperature factor entry in the PDB file is to be used as the Dmax value. The distance of any restraining atom, when rotated into the density, must be less than Dmax from the approprite target position. Line 15c path to PDB file of target position corresponding to every restraining atom. The x,y,z positions are the sites in the map to be used as targets. Every line in the "target" file corresponds to a line in the "restrain" file. BUT extra line in the target line correspond to sites that must be avoided, such as the carbohydrate site 318 on E2 when fitting E1. A sphere of rdif is drawn about these sites. Any fitted E1 atom that falls within this sphere gets counted as being "NEAR". See the comments on weights below. line starting with PIXL can determine the pixel size used to determine target position.Here the xterm value is set to the "standard" pixel size. Then all target position coordinates are muliplied by base=a1/standard. If no value for pixel is given as the first line of the file then base=1.0 This option is useful when checking on the correct pixel size.

/bio/indiana/mgr/EMmaps/corr_symm_219_368_sc.em /u19/mr/RRvfit/E1.1/mono.pdb /u19/mr/RRvfit/E1.3a/mono_restrn_tight.pdb /u19/mr/RRvfit/E1.3a/mono_target+318.pdb /pucc/scratch6/mr/junk-1.pdb /pucc/scratch6/mr/junk-2.pdb /pucc/scratch6/mr/junk-3.pdb /pucc/scratch6/mr/junk-4.pdb /pucc/scratch6/mr/junk-5.pdb end ~

Setting up the Control File (10) (NCS) Line 17 ncryst, crit2, rdif ncvryst is the number of NCS operators Crit2 defines modified fit criterioin = Fitsum –crit2*nclash Rdif (in Å.) used to compute nclash

/bio/indiana/mgr/EMmaps/corr_symm_219_368_sc.em /u19/mr/RRvfit/E1.1/mono.pdb /u19/mr/RRvfit/E1.3a/mono_restrn_tight.pdb /u19/mr/RRvfit/E1.3a/mono_target+318.pdb /pucc/scratch6/mr/junk-1.pdb /pucc/scratch6/mr/junk-2.pdb /pucc/scratch6/mr/junk-3.pdb /pucc/scratch6/mr/junk-4.pdb /pucc/scratch6/mr/junk-5.pdb end ~

Setting up the Control File (11) (NCS) Lines 18 (ncryst times) to define NCS Capa(n), psi(n), fi(n), nopt(n,1), nopt(n,2) Polar coordinates (Rossmann & Blow 1962 definition) of nth sequential NCS operator nopt defines the sequence of operations (see manual) to generate quasi symmetry.

/bio/indiana/mgr/EMmaps/corr_symm_219_368_sc.em /u19/mr/RRvfit/E1.1/mono.pdb /u19/mr/RRvfit/E1.3a/mono_restrn_tight.pdb /u19/mr/RRvfit/E1.3a/mono_target+318.pdb /pucc/scratch6/mr/junk-1.pdb /pucc/scratch6/mr/junk-2.pdb /pucc/scratch6/mr/junk-3.pdb /pucc/scratch6/mr/junk-4.pdb /pucc/scratch6/mr/junk-5.pdb end ~

Line 13: STHETA1,STHETA2,STHETA3,DTHETA1,DTHETA2,DTHETA3, FTHETA1,FTHETA2.FTHETA3 (free) (S_tart) Intial Eulerian angles, (D_elta) Increment the angles, (F_inish) Final Eulerian angles (used only for search) these limits are used only when isearch=2. When isearch=1 the limits are hard-wired as 0<=theta1<360. in intervals of dtheta1 0<=theta2<180 in intervals of dtheta2 0<=theta3<360 in intervals of dtheta3 Line 14: (CENTER(i),i=1,3),DX,DY,DZ,(FXYZ(i),i=1,3) (free) center(i),i=1,3 Site in map 1 where the model's center of gravity is to be placed in A. relative to origin. dx,dy,dz increment in center position (center(i) +/- fxyz(i)) are the limits for the translational search or climb Line 15a path to atomic model coordinates input file (PDB) Line 15b path to PDB file of restraining atoms The temperature factor entry in the PDB file is to be used as the Dmax value. The distance of any restraining atom, when rotated into the density, must be less than Dmax from the approprite target position. Line 15c path to PDB file of target position corresponding to every restraining atom. The x,y,z positions are the sites in the map to be used as targets. Every line in the "target" file corresponds to a line in the "restrain" file. BUT extra line in the target line correspond to sites that must be avoided, such as the carbohydrate site 318 on E2 when fitting E1. A sphere of rdif is drawn about these sites. Any fitted E1 atom that falls within this sphere gets counted as being "NEAR". See the comments on weights below. line starting with PIXL can determine the pixel size used to determine target position.Here the xterm value is set to the "standard" pixel size. Then all target position coordinates are muliplied by base=a1/standard. If no value for pixel is given as the first line of the file then base=1.0 This option is useful when checking on the correct pixel size.

Lines 16 paths of atomic model coordinate output files containing the rotated, best fitted, atomic model (PDB) for the top best results. Enter as many paths on sequential lines as wanted. one path only gives output for only the top fit; two paths for the two top fits, etc. End the list with the word 'end', left adjusted, on the next line after the last path. MODEL SYMMETRY referred to Map Coordinate system Line 17: NCRYST,WHTS,RDIF,NCS (free) Number of crystallographic + NCS symmetry operators defined in the h cell If ncryst=0 then assume no further input WHTS is a switch to determine whether to read in (Line 20) non default weights used as fitting restraints..eq.0 do not read;.ne.0 do read RDIF: CA atoms in symmetry related subunits that are less than rdif A. from a Ca atom in the ref molecule are noted on the output. The total number of such atoms (NCLASH) are used as a restraint in the fitting process. NCS: the number of NCS matrix operators are to be read normally ncs=0. This is only used when the NCS operators are known in matrix form instead of the polar angles supplied in line 18a. Theese matrecies are read in line 18b.

NOTE on NCS operators: There are two ways of feeding NCS operators to the program, represented by input on lines 18a and 18b. These are not exclusive, both can be used in the same job. HOWEVER if 18b is used (ncs.gt.0) then the pgm should not be used for climbing as the center position is assumed to be the input center (NOT the newly refined center. This is to allow the setting up of difference maps with previously determined NCS operators.:wq In line 18a the operators are given as polar angles (capa,psi,fi) that define a rotation axis passing through a point defined by skew values. If 360 is devisible by capa it is assumed that the operator is "proper", that is it operates 360/capa times. Otherwise it is applied only once. The operators are applied sequentially. Thus x''=[R2]x' and x'=[R1]x giving x''=[R2][R1]x. Thus, if the operators are all proper, then this produces a closed point group such as 532 (an icosahedron). Care needs to be exercised as to the result when the operators are improper. In line 18b (optional- only if ncs>0) operators [O(n)] can be supplied in matrix form. This is useful if a prior fitting operation of independent molecules generates an improper rotatio + translation operator. No check is made that the rotational component is a true rotation. The sequence of operations on NCS operators: is x(m)=[O1]x(1) and uses as many [O] matrecies as there are input matrecies then the polar operators are read. These operate on each other if they are "proper" (360/kappa) is an integer. Otherwise they operate only once. thus x(n)=[R(n)]x(m) or x(n)=[R(n)][R(n-1)]...[R(1)][O(m)]x(1) However if the original coordinates were in system y then x(1)=[E]y where [E] is the Eulerian rotation operator Hence x(n)=[R(n)][R(n-1)]...[R(1)][O(m)][E]y

Lines 18a: CAPA(n),PSI(n),FI(n),nopt(n,1),nopt(n,2),skew(n,1),skew(n,2) (free) ncryst lines, each giving another non-cryst operator. Polar coordinates in degrees using the Rossmann & Blow definition NOTE: these operators are defined with respect to the orthogonal p cell. nopt(n,1), nopt(n,2) are the first and last operators of the resultant list on which this operator is to work. The first operator is a unit matrix. Thus the first operator will operate on the unit matrix (360/capa(1)) times. The second operator will take the list of matrecies produced by the first operator and operate (360/capa(2)) times on each of the matrecies starting with matrix nopt(2,1) and finishing with matrix nopt(2,2). Etc for the third operator if it exists. skew(n,1),skew(n,2) define the x,y, intersection of the rotation axis with the z=0.0 plane. Thus if the axis is radial (as is an icosahedral axis) then these two parameters are 0.0, 0.0. Line 18b only if ncs>0 Lines 18b: (((OPERATOR(n,i,j),j=1,4),i=1,3),n=1,ncs) (free) matrix input describing a NCS operator. This matrix relates the reference molecule to each of the NCS related other molecules to be fitted simultaneously into density. Line 19a: (LFTEST(l),l=1,lf) (free) Only use the symmetry operators that are non zero for checking fit of atoms into density. I could check all (60 say) operators, but that would be a waste of time if the map had the symmetry of the operators. With this test I need test only quasi symmetry related subunits for their fit to density. Forinstance if there are the symmetry operators (defines the icosahedral 5-fold) (defines the quasi 2-fold for T=4) (defines the quasi 6-fold for T=4 corresponding to the icos. 3-fold) These operations will produce a total of 11 matrecies. Of these we can select the 1st, 6th, 7th and 8th as the matrecies that will generate the four, quasi T=4 related subunits. Hence the vector LFTEST would be NOTE: Any subunit designated as -1 will be entirely neglected. Any subunit designated as 0 will be considered as part of the background atoms. Thus clashes are calculated between background and molecule 1 between background and molecule 2 etc as well as between molecule 1 and any other non zero molecule as well as between molecule 2 and any other non zero molecule etc.

Line 19b: (TREF(i),i=1,lf) Give NCS symmetry number. For instance if T=4 there are 4 different NCS related molecules. All the others are related by crystallographic symmetry to one of the 4 independent NCS molecules. The value of the temperature factor in the O/P pdb file will be set to ((r * 100.0)/biggest) where are is the density at that atom. For the example above TREF would be Line 20: (WHT(i),i=1,6) (free) Read only if WHTS on line 17.ne.0. these weights are applied to the restraints on : wht(1) maximizing SUMF, the fit into density wht(2) minimizing the number of atoms CLASHing between subunits wht(3) minimizing the number of atoms in density < average height wht(4) minimizing RMSDST of restraining to target atoms (not used) wht(5) minimizing AVGDST of restraining to target atoms. wht(6) minimizing NEAR, the number of atoms within rdif A. to another structure to be avoided Line 21: (only if isearch=3) N,Spsi,Fpsi,Dpsi,Sphi,Fphi,Dphi (free) N gives the nth NCS element that is to be explored Spsi, Fpsi, Dpsi psi is to be explored from Spsi to Fpsi in incs of Dpsi Sphi, Fphi, Dphi phi is to be explored from Spsi to Fphi in incs of Dphi

BACKGROUND Interpolation of density in map2: let a grid point in map1 be at (ijk1(i),i=1,3) let a grid point in map2 be at (ijk2(i),i=1,3) Hence, since equivalent grid points are the same distance from the virus center in both maps, (ijk1(i)-orig1(i))*a1= (ijk2(i)-orig2(i))*a2 Therfore ijk2(i)=(((ijk1(i)-orig1(i))*a1)/a2)+orig2(i) * FOOTNOTE regarding iverbose output if iverbose.ge then output also atomap.f,maplot.f,maplot2.f output a map with numbers for density and letters for atoms in the -x,-y,z related position A=1 atom at grid point, B=2 atoms, etc. climb.f cycle number and current parameters search criterion for each trial horizscl.f a1,corr ascl,bscl,fact,nr out3d.f description of map output section kz (when kz.ge.0) map O/P (when kz.ge.0) hmin,hmax (when kz.ge.0) search.f search criteria in sections (entry before climb, set isearch.ne.0) search.f search criteria in sections (entry after climb, set iclimb.eq.2) stormap.f section kz hmin,hmax average,rmsden,pts,rr1,rr2 vertscl.f rin,rout,corr,ascl,bscl,rfact,rmsdiff,rmsr1,rmsr2,nr Disc units: unit=1 first input map =2 second input map =3 third input map that provides guidance for mask construction =4 output: scaled (input1 - input2) maps =5 input control file =6 output log file =7 atom PDB input =8 atom PDB restraint file =9 atom PDB target file =10,11,...19 PDB output files for top 10 solutions

Output information Echo of input Analysis of input map(s) for density distribution with radius NCS matrices ntop search results Climb results List of atomic clashes Analysis of number of atoms in density bins Maps of density and of atomic positions

CLOCKIT 04/25/10 11:28:45 CPU time (sec):.030 User:.010 Syst em:.020 Page faults: 0 map 1 input file is called /bio/indiana/mgr/EMmaps/corr_symm_219_368_sc.em Purdue EM format map 2 is not in use map 3 is not in use map 4 is not in use first input EM map is scaled by secnd input EM map is scaled by third input EM map is scaled by distance a1 between pixels in first map increment a1 by da1 in na1 steps distance a2 between pixels in secnd map distance a3 between pixels in secnd map origin pixels in first map origin pixels in secnd map origin pixels in third map move origin flags for the three maps turn maps through 90deg ? (0/1 no/yes) ext. int. & inc for map scale determ hand of maps number of variables for scale factor 2 set all density to zero beyond 400.0A. density within radrna not in mask is rna use only (rho2-ave_rho2)>crit1*rms(rho2) 3.00 verbosity of output 2 apply local vertical scaling: 0=n, 1=y 0 ismear,jsmear for mask making with map3 0 0 maptype (0=diff; 1=map1; 2=map2) 1 first and last section output print icosahedral triangle if icos.ne.0 1 iwrstr: controls pre-superpositions 0

ANALYSIS of MAP 1 pixel size , map center rad in rad out pts ave den rms den hmin hmax biggest absolute value on stored map is The number of pixels with this value is 33955

density distribution in map 1 rr1 rr x rms density

map density section

the top 25 points in search id theta1 theta2 theta3 centx centy centz sumf clash -den avgdst near

variation in fitting restraints after general search sumf clash -den rmsdst avgdst near average sigma weights variation in fitting restraints after climb & before contraction sumf clash -den rmsdst avgdst near average sigma weights CLOCKIT 04/25/10 11:55:10 CPU time (sec): User: Syst em: Page faults: 2 possible number of fits found after climb 25 by removing similar results this contracted to 7 # rcrit value sumf clash -den avgds near thet1 thet2 thet3 centx centy centz rcrit= sum [(x - mean x>/(rms from mean x)] over all criteria CLOCKIT 04/25/10 11:55:10 CPU time (sec): User: Syst em: Page faults: 2 fit # 1 distribution of atoms in density heights domain domain domain

atom distribution section map density section

/bio/indiana/mgr/EMmaps/corr_symm_219_368_sc.em /u19/mr/RRvfit/E1.1/mono.pdb /u19/mr/RRvfit/E1.3a/mono_restrn_tight.pdb /u19/mr/RRvfit/E1.3a/mono_target+318.pdb /pucc/scratch6/mr/junk-1.pdb /pucc/scratch6/mr/junk-2.pdb /pucc/scratch6/mr/junk-3.pdb /pucc/scratch6/mr/junk-4.pdb /pucc/scratch6/mr/junk-5.pdb end ~