SedAnal v5.54 2010-4-1

Known bugs: The "Check grid" alert doesn't appear immediately sometimes when the errors are occurring continuously; this will be fixed in a future release.

 

Specific changes:

 

1) The limits are now enforced when doing F-statistics. Previously, they were ignored.

SedAnal v5.53 2010-3-13

Known bugs: The "Check grid" alert doesn't appear immediately sometimes when the errors are occurring continuously; this will be fixed in a future release.

 

Specific changes:

 

1) Simulation of equilibrium data has some corrections:

a) The cell data file (.abr) is flagged as an equilibrium run, not velocity.

b) In the case of simulation when both a cell data file and a prototype control file are produced, and when the simulation control file has blanks in its name, the cell data file is now correctly named. Previously, the blanks in the cell data file's name were replaced with "!".

c) The prototype control file includes an initial guess of 0.0 for yOffset. Previously, this was left blank. As before, the parameter is fitted.

 

SedAnal v5.52 2010-3-5

Known bugs: The "Check grid" alert doesn't appear immediately sometimes when the errors are occurring continuously; this will be fixed in a future release.

 

Specific changes:

 

1) The program has been rebuilt so no vcomp.dll is required. This caused an immediate error "application configuration is incorrect" if you didn't have the exact version of vcomp.dll used in the SA build.

 

SedAnal v5.51 2010-2-22

Known bugs: The "Check grid" alert doesn't appear immediately sometimes when the errors are occurring continuously; this will be fixed in a future release.

 

Specific changes:

 

1) When starting SA, if a preferences file did not exist, and user said no to creating one, SA crashed, because it was attempting to write a session file in the wrong directory. This has been corrected.

 

2) For equilibrium fits for the special case monomer-dimer, the function being fit was incorrect, which meant results for this special case could have been incorrect. This has been corrected.

 

3) For 3rd virial coeff (CM1) in equilibrium fits for the special case monomer-dimer, the Jacobian was incorrect, which meant fits could have taken longer than necessary. This has been corrected.

 

4) The equilibrium fitter (simulator) now handles BM1 better. For a single species or monomer-dimer, up to about BM1=5 works. If the simulator doesn't converge, an alert appears. Previously, no alerts appeared, and in some cases it was possible to get plausible output even though convergence hadn't occurred.

 

FYI, different algorithms are used for each of

a) Ideal case: iterate on c(r) and c(men).

b) Monomer-dimer: use binary search for c(men), making use of c2(men)=K*c1(men)^2.

c) Non-ideal case: iterate on c(r) and c(men), solving simultaneous non-linear eqns for c(r) and nonIdeal each iteration.

The non-ideal case is new. Previously, nonIdeal=BM1.c+CM1.c^2 (.=dot product) was calculated at each iteration from the previous iteration's values. Testing should include each of these cases.

 

5) The STOP button now works during equilibrium fits. With large non-ideality, these can take significant time, and may not converge in extreme cases.

 

6) The number of iterations for equilibrium fits is now shown in the C window on the fitting screen.

 

SedAnal v5.50 2010-2-8

Known bugs: The "Check grid" alert doesn't appear immediately sometimes when the errors are occurring continuously; this will be fixed in a future release.

 

Specific changes:

 

1) Scan selection in the preprocessor now works for the data scans, even if the scan numbering for blanks overlaps that for data scans.

 

2) Reloading a simulation control file no longer resets the load conc type and units to the defaults.

 

3) Under certain conditions, the Info area for blanks was disabled. It is now always enabled

 

SedAnal v5.49 2010-2-6

Known bugs: The "Check grid" alert doesn't appear immediately sometimes when the errors are occurring continuously; this will be fixed in a future release.

 

With respect to the molar limits when load conc input was g/L, the limits that were in the control file you sent were already

[This should have been included in the 5.48 release notes]

The format of cell data files (.abr) has been changed. Older files will work in v5.48, but those produced by 5.48 will not be read properly by earlier versions.

 

Specific changes:

 

1) Preprocessor now has a different interface for blanks used in equilibrium runs. The attached document describes the details.

 

2) The title at the top of the screens that shows the SedAnal version now has a 4-digit number following the version (e g, v5.49.0015). This is a build number, and you can ignore it.

 

3) The preprocessor no longer has a separate icon on the task bar. This was an old bug.

 

4) If you loaded a control file with user eqns involving load concs, then clicked New on the control screen, in some cases SA hung. This has been corrected.

 

5) On the fitting screen, the values shown for ticks or grid are now 3 sig figs, rather than 2. For short cells, 2 sf was not enough.

 

6) When an existing control file is loaded, the default limits for the empty cell data files depends on the load conc input type and units. Thus, when a cell data file is added, the limits will be these defaults. Previously, the default limits were appropriate for mol/L or mol/L ratio. The new default limits are 0-55 mol/L, 0-100 g/L, and 0-1e6 for either weight or molar ratio.v5.49

 

 

SedAnal v5.48 2010-1-28

Known bugs: The incorrect limits change when adding a dataset hasn't been fixed yet; you can work around this by just changing the limits on the control screen. I have also noted that the "Check grid" alert doesn't appear immediately sometimes when the errors are occurring continuously; this will be fixed in a future release.

 

Specific changes:

 

1) Preprocessor now has a different interface for blanks used in equilibrium runs. A separate document that will be included with v5.49 describes the details. This feature isn't finished yet.

 

2) Blank-related bugs fixed:

a) Radial interpolation for blank could have been off by 10^-4 cm.

b) Description and run date/time stored in cell data file were those of the last blank added, not the data file. For existing cell data files, the incorrect description can be changed using the preprocessor, but the run data/time cannot be changed.

c) In certain cases, rescaling the plot gave an incorrect (large) yMax and yMin, so the plot "disappeared".

3) For multi-component systems where the second component is input in absolute units (e g, g/L) and the loading concentrations of the first component are constrained by user equations, changing the second component could cause screen updates that were incorrect. This has been fixed.

 

4) The Package creator may have written zip files that were not correct; this has been fixed.

 

SedAnal v5.47 2009-12-29

1) Preprocessor no longer crashes after getting an error due to reading an invalid cell data file.

 

2) Centrifuge scan files are now loaded properly in the preprocessor; v5.45 and 5.46 didn't load them.

 

SedAnal v5.46 2009-12-18

1) The experiment folder list is now populated in the Preprocessor.

 

 

SedAnal v5.45 2009-12-18

1) Packages (for fitting or simulation, but not dc/dt) can now be imported. Importing copies the contents of a package (the control file, cell data files, and model file) onto your machine. Details are in the attached document.

 

2) The max and min were reversed when intensity data was rescaled. That caused the plot to be inverted, and no grid lines would show. This has been corrected.

 

 

SedAnal v5.44 2009-12-7

1) The load conc limits are now converted when the units and/or input type (abs or ratio) is changed. This functionality was not complete in v5.43, contrary to the release notes.

 

2) The dc/dt weight average window

a) Can be copied and pasted (^A selects all, ^C copies to clipboard; you can also select individual lines).

b) Has units for all quantities.

c) For WDA, only Sw, Cp and Co are shown.

d) For WDA, Sw is correct for ranges less than the complete peak.

e) If the user selects a narrow range (not normal) that includes <2 points, SA selects 2 close by, if possible; otherwise it gives an error message.

3) An error prevented the correct value for omega^2*t from being used in the "g of S" report in dc/dt. This error was introduced in v5.36, and has been corrected.

 

4) For scripts

a) If a script file was selected from the dropdown list of recent scripts, but the file had been deleted or moved, an error message appeared, but SA then crashed. Now there is no crash.

b) In a script, if a line of the form "LOAD \experiment\controlFile" appears, the user gets a warning about the illegal initial \, and it is removed.

5) In the proprocessor

a) Writing adjusted data file (i e .abr) turns off the type of run, as previously. Now this also disables the "add blanks", so loading a new scan file includes only the scans in that file.

b) Usage note: to add scans to an existing .abr file of an equilibrium run, load the .abr file, then click Equilibrium (that choice will already be set). Then add blank files as usual. This feature is unchanged from previous versions.

6) On the control screen, the eqn editor now allows the cell number (initial subscript for extinct coeff) to be 0, meaning all cells.

 

 

SedAnal v5.43 2009-11-23

1) Preprocessor re-creating scan files:

a) Writing a set of scan files from a cell data (.abr) file in the preprocessor produces output with 2 or 3 columns, radius, optical density (OD), and, if available, std dev of OD.

Previously, the std dev column was -1.0 for deleted points, and missing for other points, even if there was std dev data in the original scan files.

Now, the std dev column contains the std dev that was in the original scan files. If there was no 3rd column in the original scan file (e g, if it is interference data), no 3rd column will appear in the recreated scan files.

The std dev information has always been stored in the cell data files.

b) The radius is being output with one fewer blank preceding it; it is now "bbb5.7980bbb-1.53000E-01bbbb2.61000E-02" (b=blank).

2) The "check grid" message has been made more prominent, and an explanation appears when you mouse over it:

 

 

 

3) The load conc limits are now converted when the units and/or input type (abs or ratio) is changed.

 

SedAnal v5.42 2009-11-17

1) The dynamic density correction for sedimentation coeff (currently a developer option) has been changed to

densityCorr = 1.0 - 1e-3*DOT_PRODUCT(vBar(1:nSpecies,iRun),Cm(iRadius,1:nSpecies))

Previously, the sign was +.

 

SedAnal v5.41 2009-11-17

1) On the control screen, clicking the yellow or green small square to switch between ratio and conc input did not do the calculation correctly, so the converted value was not the same as the original. This has been fixed, so now all the conversions (w<->c, w/w<->c/c, w<->w/w, c<->c/c) should give equivalent values.

 

2) A crash when doing certain simulations has been fixed. This crash occurred because of stack overflow.

 

SedAnal v5.40 2009-11-13

1) In dcdt

a) dcdt0 is specified by the user as an increment. In other words, the value the user chooses is added to the previous value of dcdt0. While the user is holding the left mouse button down and adjusting the black horizontal line, the "dc/dt zero at box" and line will show the change in dcdt0 to be added. When the user releases the mouse button, the box and the plotted red line will show the value of dcdt0.

 

For example

Load data: User loads data. dcdt0=0, no red line, box empty.

Adjust dcdt0: User clicks "dc/dt zero adjust" button, positions mouse in dc/dt plot at dc/dt (y-axis)=-0.02 and holds down left mouse button. Black line appears at -0.02 and box shows -0.02.

Adjustment done: User releases left mouse button. Black line at -0.02 changes to red, and box continues to show -0.02. dcdt0=-0.02.

 

Re-adjust dcdt0: User clicks "dc/dt zero adjust" button again, positions mouse in dc/dt plot at dc/dt (y-axis)=+0.005 and holds down left mouse button. Black line appears at +0.005 and box shows +0.005.

Re-adjustment done: User releases left mouse button. Black line at +0.005 disappears, and red line appears at -0.015. Box changes to show -0.015. dcdt0=-0.02+(+0.005)= -0.015.

 

b) The weight averages for WDA have been changed to do a numerical integration rather than assuming a fixed grid. Sn, Sz and Sz1 are not displayed. The weight average window for WDA now shows only Sw, with no error limits.

c) The average report for WDA was not using the last scan. Now it is.

d) The average report for WDA now contains the 10 columns

 

2) The dynamic density correction for sedimentation coeff (currently a developer option) has been changed to

densityCorr = 1.0 - 1e-3*DOT_PRODUCT(vBar(1:nSpecies,iRun),Cm(iRadius,1:nSpecies))

Previously, the sign was +.

 

3) For simulations, total scans of 1000 is the maximum that can be accomodated. Requests for more gives an error message.

 

4) On the control screen, shift-left-click brings up the limits window, even if the parameter is being held constant. Limits are not shown for derived parameters, as before. This is true for both the simulation and analysis screens.

 

5) The choice of weight or molar units for load conc is preserved when loading a simulation control file. This error was introduced in v5.38 by my dog.

 

6) On the control screen for simulations, the tool tip for the load conc header now works the same as for analyze: it tells you which color square to click to change units or switch between ratio and conc input.

 

SedAnal v5.39 2009-11-2

1) The cell description has been restored to the report.

 

2) Runs with 1000 scans could not be loaded (999 was the limit).

 

3) Preferences / dc/dt no longer crashes when invoked.

 

4) WDA with lower limit of range for averages ln s*<0 now computes wt avg.

 

5) The names of variables xFitRange0 and xFitRange1 have been changed to xAvgRange0 and xAvgRange1 in dc/dt control files. This means that control files created (using package) prior to v5.39 when read will not restore the range for averages, and also will give a warning that the lines with xFitRanges are ignored.

 

6) If repeated fits were done without exiting SedAnal, the list shown in the I window had its top and bottom cut off. This has been fixed.

 

SedAnal v5.38 2009-10-26

1) The calculation of D in the non-ideal case for sedimentation velocity fitting has been changed; the coefficient of c^2 has been changed from 3/2 to 3. In v4.75, (old) was added as a developer option; in v5.06, it was made the default.

nonIdeal=(1.+DOT_PRODUCT(Ks(1:nSpecies),Cm(iRadius,1:nSpecies))        (old)

         +DOT_PRODUCT(Ks2(1:nSpecies),Cm(iRadius,1:nSpecies)**2))/

     (1.+2.*DOT_PRODUCT(BM1(1:nSpecies),Cm(iRadius,1:nSpecies))

         + (3./2.)*DOT_PRODUCT(CM1(1:nSpecies),Cm(iRadius,1:nSpecies)**2))

nonIdeal=(1.+DOT_PRODUCT(Ks(1:nSpecies),Cm(iRadius,1:nSpecies))        (new)

         +DOT_PRODUCT(Ks2(1:nSpecies),Cm(iRadius,1:nSpecies)**2))/

         (1.+2.*DOT_PRODUCT(BM1(1:nSpecies),Cm(iRadius,1:nSpecies))

         + 3.*DOT_PRODUCT(CM1(1:nSpecies),Cm(iRadius,1:nSpecies)**2))

In either case, DiffCoeff=DiffCoeff/nonIdeal. Concentrations in these equations are in mass units.

 

2) In dc/dt

a) A warning has been added when sCorrection is too close to the singularity. If v-bar*density is within 0.001 of 1, a warning is given, and correction will not be applied.

b) The "Compute weight averages" check box is gone. Averages will always be calculated. If no range for averages has been chosen, the averages window shows "Select a range for averages". The range is selected separately for dc/dt and WDA. Also, when switching between normal and WDA, or loading new data, any range already chosen is removed.

This makes the corresponding value in Preferences obsolete. However, the user had no way to change that (it was not on the dc/dt preference screen).

c) Both dc/dt and the average dc/dt now have dcdt0 subtracted, so they track when the zero is changed. Previously, only the average was adjusted.

 

d) When switching among which of dc/dt data and avg to show, rescaling does not take place. This is the only way to preserve the zoom state. In the future I will add a "rescale" button.

3) A new option controls which scan pairs are printed in the minimum files. The default for existing control files is "All". The attached document describes how to specify the files in Preferences or in a control file.

 

4) On the control screen, changing from velocity to equilibrium uses the model defaults for the parameters. These include the limits, which were not being changed previously.

 

SedAnal v5.37 2009-10-19

This version should have almost all the dc/dt fixes, but the scan pair select is not complete. The user interface is present, but it doesn't affect min files yet. I wanted to get this to you so you can try the dc/dt changes.

 

One dc/dt issue to be resolved is that if Compute weight averages is checked, but a range hasn't been selected,  garbage values are displayed. I'm not sure what to use as a default range.

 

The error limits are all being shown as 0, because we don't have an estimate for stdDevMean. When you have determined that, I will add it.

 

Specific changes:

 

1) A simulated equilibrium run could have crashed for certain data. This has been fixed.

 

2) In dc/dt

a) The Adv settings now show the defaults when no data has been loaded.

b) The range of s* is now changed when the viscosity ratio, density, or v-bar is changed. Previously the range was 0 - (1/(w2Final*tSed))*log(7.3/Rm); now it is 0 - (1/(w2Final*tSed))*log(7.3/Rm)*sCorrection, where sCorrection = viscRatio*(1-vBar20*density20)/(1-vBar*density).

c) Change in viscosity ratio, density, or v-bar forces rescaling of both plots.

d) A bug in the display of dcdt0 has been fixed. This displayed the red line in the wrong place if zoom was followed by unzoom. The correct value was still used in calculations.

e) For WDA, the value of s* is divided by sCorrection (defined in (b) above) after the plateau correction, instead of before.

3) The calculation of s during fitting has a developer option "Dynamic density correction for s". If this is checked, s will be multiplied by the factor

 

V=1.+1e-3*DOT_PRODUCT(vBar(1:nSpecies)*Cm(iRadius,1:nSpecies))

 

where vBar(iSpecies)=(1-densIncr(iSpecies))/density(iSpecies). The value of densIncr is taken from the molecular parameter, and density is from the experiment data.

 

Note that the value of vBar in experiment data will not be used. Also, this may or may not work for control files in which psv is used instead of density increment, or which use M, D and s as the parameters.

 

4) An initial F-stat increment has been added for D.

 

5) A crash occurred on certain fits (e g, non-ideal 2A=A2) for simplex fitting. This has been fixed.

 

6) The preference values for min files were not being used as defaults for new control files, or old control files that didn't have these values. In addition, the "save dc/dt" variables were not being stored in the preference file at all. These omissions have been fixed.

 

7) There are new fields in Preferences/Control extended/Output/Minimum files and Outputs/Minimum files on the control screen. These are not finished, and will not have any effect on the minimum files, although they will store values in the control files.

 

SedAnal v5.36 2009-10-1

Specific changes:

 

1) The maximum number of points in a scan for a simulation which writes an .abr file has been increased from 5000 to 10000.

 

2) The test for using the non-ideality eqns for s and D has been changed. Now, if any species has a non-0 Ks, Ks2, BM1 or CM1, non-ideality is used for all species.

 

3) A simulated equilibrium run

a) Will now produce a run (.abr) file and a prototype control file, if these options are set in Advanced/Simulating.

b) Will output scan files only if "Output individual scan files" is checked. Previously, scan files were always produced.

4) The default density of water has been changed from 0.9982 to 0.99823; the default partial specific volume from 0.73 to 0.725.

 

5) In dc/dt, the s* vs g(s*) plots and report have been changed as follows:

sCorrection=viscRatio*(1-psv_20*density_20)/(1-psv*density);

Then s* is computed for the experimental data as

s(iRadius)=log(rGrid(iRadius)/Rm)/(w2Final*tSed) * sCorrection

The s* grid is created by using the default or user-chosen ds* and the specified number of s* grid points, then the experimental data is interpolated onto this uniform grid.

 

When the Adv button is clicked, a small window appears, showing the viscosity ratio, density, and partial specific volume (psv). Initially, the values are taken from the Experiment Information (preprocessor); if no experiment information had been entered, these default to 1, 0.99823 and 0.725, respectively. Entering new values and clicking "Apply" recalculates sCorrection, redoes the computation of g(s*) and gHat(s*), and replots the data in both windows. Clicking "Restore experiment values" returns to the initial values.

 

6) For dc/dt WDA, selecting scans now gives you access to all the speed sets. Formerly, you could just select scans from the first, and these were ignored.

 

SedAnal v5.35 2009-9-14

*WARNING* This is not a fully-functional version. In particular dc/dt will not work, and experiment information hasn't been fully tested. Use it only where v5.34 won't do the job, and don't give it out.

 

Specific changes:

 

1) There is a new Preference/Control option to make a copy of the control file for every fit. The copies are named <control file name>_nnn.abc, where nnn is the same number as used for the reports and the general log.

 

Existing control files will have this option turned off; new control files will pick up the preference setting, and cannot be changed. If it is absolutely necessary to change old control files to have this option, it can be done with a text editor (variable controlFileCopy).

 

2) The maximum number of parameters has been increased from 625 to 1493. "Parameters" includes both fixed and floating values. Fits with the maximum number of species, rxns, and runs can actually exceed this number, although it is unlikely. This is a temporary fix; a future version will allocate appropriate storage, and/or warn the user about the maximum being exceeded.

 

3) A new field has been added to the experimental information screen: partial specific volume. This is used only for dc/dt; for fitting, PSV is a molecular parameter, either per se, or in the form of density increment.

 

4) When simulating data, values <1e-99 are converted to 0. This conversion is done before noise, if any, is added.

 

 

 

 

 

8/22/2009                         SedAnal v5.34

1) The initial equilibration method was incorrectly being changed to numerical integrator when Newton-Raphson was specified, in the case of kinetic systems (rxns with user-specified rate constants, rather than equilibrium constants). N-R is now used if specified.

 

2) The general log now shows the initial concentrations (concentrations of all species after equilibration) in both g/L and mol/L. Previously, these did not appear in the log.

 

3) The general log filename is now numbered to correspond to the report and min file numbering. If the log is enabled, report and min file output are enabled, and report file numbering and min file numbering are enabled (these are all enabled/disabled under "Output files..." on the control screen), files generated by a fit might be

MyControlFile_Report053.rtf

MyControlFile_Log053.txt

MyControlFile_02Min053.txt

4) Scripts may now load control files with spaces in their path name (e g, My Scripts\MyControlFile.abc). Previously this looked for a control file "My.abc".

 

5) The SEulEx error control was incorrect, and thus produced results that were slightly (1-2%) wrong in some cases. The error control has been fixed, but with a tolerance of 1e-6, the integration is ~10x slower. Here are my findings for your ABCD case:

Method     Fractional      Loop times, runs 1-3

           (abs) accuracy

           required

SEulEx     1e-6 (1e-18)    262.3, 226.5, 200.8

SEulEx     1e-5 (1e-17)    213.1, 192.2, 167.5

SEulEx     1e-4 (1e-16)    129.6, 144.4, 129.0

SEulEx     1e-3 (1e-15)    40.3

SEulEx     1e-2 (1e-14)    28.2,33.9,35.9

Parallel

 SEulEx:   1e-2 (1e-14)    19.2,24.5,26.6

2 compute

 threads

These times all include generating the general log. Without that option, the times will be about 0.4 sec faster.

6) This version has an experimental parallel kinteic integrator (see timing above). It can be used *ONLY* for

a) SEulEx method

b) A model which has the ABCD rxn scheme (A+B=C, B+C=D)

You access this by checking two boxes on the Developer options screen. You must check both.

For SEulEx, solve ABCD directly using degree of rxn, tather than concs

Numerical integration for kinetics is executed in parallel

7) The ClavSim ("C") window on the fitting screen now shows additional information:

a) Whether the computations are being done in parallel or serially (the old way); there is a separate indicator for the kinetic integrator and the Claverie FE solution. The latter is not yet parallelized, so it will always show "serial". At the moment, the setting of the former is determined by your choice of the developer option for parallel KI ((6) above).

b) The number of processors and maximum number of threads on your machine. The term "processors" here means the number of independent calculations that can proceed simultaneously, so it includes both cores and hyperthreads.

c) The number of maximum number of threads and the number of threads in use. The latter changes as a parallel section of the pgm is entered (in this case, the KI). "Threads avail" is not now being used, and should be 0.

d) A log window shows the "loop time" for each run. This is the time required to simulate sedimentation and kinetics for the entire run. It is the same as the "Claverie loop time" which appears in the general log (if you have that log enabled).

 

 

7/14/2009                         SedAnal v5.33

1) For dc/dt WDA, reloading the data resets which buttons are pressed initially to 6.5, and shows a single plot. In v5.32, the single plot for 6.5 was also shown, but other buttons were shown as pressed.

 

6/27/2009                         SedAnal v5.32

1) Simulations of systems with more than one species failed, due to an error introduced in v5.30 in connection with the variable grid spacing. This has been fixed.

 

6/26/2009                         SedAnal v5.31

1) When dc/dt zero level is changed, the display of g(s*) (or log s*xg(s*)) in the right window changes as follows:

a) If the right window is unzoomed, the data will be replotted, the entire range of s* will be shown, and the y-axis will be rescaled.

b) If the right window is currently zoomed, the data will be replotted but not rescaled. The user can type ^Z one or more times to return to the unzoomed state, and after that the entire range of s* will be visible.

 

6/25/2009                         SedAnal v5.30

1) The progress message on the fitting screen and the header in the fitting report have been corrected to include F-statistics. Previously, either "bootstrap (subset)" or "Monte Carlo" appeared.

 

2) The control screen "Output files.../Minima" page and the "Preferences/Control extended/Outputs/Minima" pages have had the descriptions reworded.

 

3) For derived parameters calculated from an eqn:

a) The report has been modified so the value used for the best fit is reported. The derived parameters are still listed as having fixed values, with a note showing how they were computed. For example,

Constraints on parameters

   Source     Equation

...

16 user         e(4,1)=e(3,1)

 

Parameters

Name                                               Fixed value       Initial value     Best fit           Limits (low - high)

Molecular mass (g/mole) of A                                      4.46560·104    4.46560·104    0.00000 - 1.00000·1020 

...

Mass extinction coeff. of A in run 1    5.84400·10-1 

Mass extinction coeff. of A in run 4    9.00788·103 (calculated from constraint 16)

 

b) The value used for the best fit appears on the control screen when Last is double-clicked, or B is clicked, to show the best-fit values.

4) When doing F-statistics, a pair of developer options allows you to do only the + or only the - side for all parameters selected for F-stats. If you are doing F-stats on n parameters, you can run the data on 2n machines, using one machine for + and one for - for each parameter.

 

4/28/2009                         SedAnal v5.29

1) The selected range was not being used properly for either the weight average or the report. This has been fixed.

 

4/13/2009                         SedAnal v5.28

1) When the fitting window is resized, the tooltip now moves with the STOP button.

 

2) When an image was saved from the fitting window, the tooltip changed to show the image filename, but did not change back. That has been corrected, as well as another potential bug for the image filename tooltip.

 

4/8/2009                           SedAnal v5.27

1) The dc/dt Preference page now checks the number of radial points against the correct limit of 3200, rather than the previous incorrect limit of 1000.

 

2) There are a couple of minor changes to the fit window tooltip handling.

 

4/7/2009                           SedAnal v5.26

1) Changes in fitting screen:

a) Equilibrium fits do not hang when x-ticks or -grid is turned on.

b) Fitting screen is maximized or not according to control file choice. Previously, Preference choice was used.

c) Tooltips and toolbar can be removed (see (4)).

2) If a fit with F-statistics was started, and interrupted by the user, the confidence limits that appeared in the report were garbage. Now, the report will say "[not completed]".

 

3) In the F-statistics log, successive sessions are separated by a line of "-".

 

4) Three new developer options turn off the tooltips, toolbar, and custom cursor on the fitting screen. These are for testing of the disappearing problem. These three are the functions associated with the mouse.

 

You have to restart SA each time you change one of these for them to work correctly.

 

 

3/21/2009                         SedAnal v5.25

1) Changes in dc/dt:

a) WDA now scales properly.

 

3/17/2009                         SedAnal v5.23

1) Changes in dc/dt:

a) A new button changes the plot of dc/dt. Choices are individual scans only, average only, or both. Clicking the button cycles through these choices. The default choice is set in Preferences.

b) The average dc/dt is now optionally smoothed, and the smoothed dcdtAvg(s*) is used to compute ghat(s*). Typing Enter after changing the smoothing percent causes a recalculation and replot of g(s*) (or s* g(s*), if log was selected).

Smoothing is done if the smooth %*nPointsis between 2 and nPoints. The special case of 100% smoothing and nPoints even is treated as nPoints-1.

c) A spurious point (0,0) on the s* x g(s*) vs ln s* plot has been removed.

d) The scaling of the plots has changed slightly; see the attached for details.

e) The tooltip delay time has been increased, so moving the cursor over a control doesn't trigger the tooltip immediately.

f) When a range for averages is selected, the g(s*) plot is shown in red for points in the range, and light gray for those outside the range.

g) The point size for the log plot has been reduced. Log plot also shows the points outside the selected range in gray.

h) A button had been added to show g(s), ghat(s), or both. This is experimental; if you decide to keep it, I will fix the labeling and scaling. Right now the y-axis label is always g(s*) or s*.g(s*), and the scaling is sometimes incorrect, cutting off the top of the curve.

You can reset the scaling by positioning the mouse in the smooth Percent box and typing Enter (without changing the value). This works in general.

The g(s*) curve is green. gshat is red, and s* x gshat is blue (as in previous versions).

i) The WDA report has been changed. The average file now has 5 columns

 

s*    ln(s*)    avg s*gshat(s*)    avg s*gshat(s*)/normFactor   avg s*gshat(s*)*ln(10)

 

with normFactor=INT(s*gshat(s*) ds*/s*) for all s*>0.

 

2) Changes in Preferences:

a) A new dc/dt preference has been added: which dc/dt vs g(s*) plots will be shown by default when a cell data file is loaded.

b) The size of the s* grid is limited to 100-1000. Previously, the user could enter any limits, and, if outside the range 3-1000, could have caused a crash.

c) The size of the radial grid is now limited to 100-3200, although higher limits will not crash.

 

3) Changes in preprocessor:

a) Selecting an .abr file which had no previous jitter adjust, and typing in the "Jitter adjust at" box showed a vertical yellow line, rather than a stripe. This now works correctly.

4) Changes in control screen:

a) v5.22 introduced a bug: you could not create a new control file and use it for fitting; you got an error that maximum iterations was blank. That has been fixed.

b) You no longer get a "limits suspicious" warning for a negative extinction coeff limit (unless both limits are equal).

 

 

3/7/2009                           SedAnal v5.22

1) Changes in dc/dt:

a) When changing ds or smoothing % by typing a value in the box, followed by Enter, the range to fit was removed. This has been corrected.

b) The "Last" list is now implemented. Selecting a file from the list, or clicking Last, restores the screen to where it was when dc/dt last exited.

c) A dc/dt control file is created when user exits with OK. The name of this dc/dt control file is <cell data>.dbc, where <cell data> is the part of the name of the .abr file excluding the date. For example, if the cell data file is 20090302_ABCD.abr, the control file will be ABCD.dbc. In the current version, there is no way to change this name, and previous dc/dt control files for the same cell dat will be overwritten.

d) The spurious points at (0,0) in WDA have been removed.

e) A bug in the zooming for WDA has been corrected. This bug occurred if a zoom area was selected, then ^Z was used to go back to the previous scaling.

f) The Package option is now implemented. This creates a zip file containing the dc/dt control file and the cell data file.

g) Smoothing is now done for the dc/dt vs s* plot. To change the amount of smoothing from the default (0), enter a number in the "Smoothing/Percent" box (the same as for WDA) and type Enter. The data will be replotted with the smoothed average (the black line in the left plot).

h) A new "Floatation" check box inverts the signs of s* and reverses the scans. At the moment, this box is only effective if it is checked at the time the data is loaded. User gets a notice to this effective when the box is checked after data has been loaded.

i) For cell data files from simulations, placing the cursor on the "Cell data" box caused a crash. This has been fixed.

2) The accuracy of deducing the control file path when only part of the path is given has been improved. This affects scripts. The complete rules are

a) If there is no path separator (a "\"), the input is a control file name. The path is completed by prefixing, in priority order,

 i) The directory of the most recent control file for this instance (session) of SedAnal (e g, Coconut becomes C:\SedAnal\User_data\MyExperiment\Coconut .abc).

ii) The directory of the most recent control file in the "Last" list for any instance (session).

b) Else if the input begins with "\\", it is assumed to be a complete network path (e g, \\MyOtherComputer\IsAGoldenDelicious\Macaroons\Sugar.abc).

c) Else if the start of the input is NOT the same as the start of the default user data path, the input is assumed to be of the form <experiment folder>\<control file> in the default user data directory (e g, MyExperiment\EggWhite becomes C:\SedAnal\User_data\MyExperiment\EggWhite.abc).

d) Else the input is assumed to be a complete path (e g, C:\SedAnal\User_data\MyStuff\Macaroons\Almond\Paste.abc).

In all cases, the ".abc" extension will be added to the input if it isn't present.

 

3) Changes on the control screen

a) The tooltip for the control filename has been moved so the filename is still visible.

b) The "Maximum iterations" box has been removed (it is now in Preferences/Control extended/Fitting).

4) Changes in Preferences

a) Maximum iterations has been added, with a default of 100,000. This applies to SV and equilibrium fits.

 

5) The audible alerts have been changed from C below middle C to middle C (256 Hz). You should now hear them even if your sound system is off (there are still a few that require the sound system be on).

 

6) Changes in the preprocessor

a) Jitter adjust button now changes from "Skip" to "Reset" when jitter adjust is done. The operations performed are the same.

b) Attempting to do jitter adjust when no data is loaded no longer crashes. The same is true for locating the integral fringe reference radius.

c) The tooltip which appears when a cell data file is written now disappears when another cell data or set of scan files is loaded. It also disappears when its X is clicked (this is the same as in previous versions).

 

 

2/24/2009                         SedAnal v5.21

1) The form of the session log has been changed to accomodate multiple sessions.

a) The session log in User_data is now named SedAnalSessionFilenn.txt. nn is the session number mod 10 (i e, 00-09). The most recent 10 session files are kept, after which they are overwritten.

b) The previous session file, SedAnalSessionFile.txt, will be deleted the first time v5.21 is run. If earlier versions are run, SedAnalSessionFile.txt will be recreated. You should not run v5.21 and any earlier version simultaneously on the same machine, as this might cause the earlier version to crash.

c) Each session has a unique ID for the machine on which it is run. This is governed by a new file, SessionID.txt, in User_data. At present, if the User_data directory is changed, the sessions will be renumbered from 1. This should not cause any difficulty, as the session files will also be in the same new directory.

d) A new file, SessionList.txt, in User_data, shows the start and end time of each session. This file will be used to diagnose problems such as the disappearing.

2) The format of the recent control files file, RecentControlFiles.txt, has changed. v5.21 will convert the older format, but once the conversion has been done, earlier versions will not read the file correctly.

 

3) Scripts now determine the directory of the control file for a LOAD command as follows:

a) If the argument of the LOAD is a complete path, it is used.

b) Else, the most recent directory of a control file stored by the current session is used.

c) Else, the most recent directory of a control file stored by any session is used.

4) Some minor UI errors in the script selection window were corrected.

 

5) The single output file for WDA, yyyymmdd_<name>_WDA.txt, has been split into two output files. The first, which retains the name yyyymmdd_<name>_WDA.txt, contains the header information and the s* vs s*g(s*) tables for the selected radii; the second file, yyyymmdd_<name>_WDA-avg.txt, contains the same header information, and the averaged values for ln(s*) vs s*(g(s*), interpolated onto a uniform grid.

 

6) The averages referenced above were off by one x-value, and the final value was garbage. This has been corrected.

 

7) There are a few changes to the dc/dt screen:

a) New buttons for "Last" appear next to the "Cell data" list.

b) A new "Package..." button appears at the right. This is not yet completely implemented.

c) All the controls now have tooltips.

d) The "Interpolate" box is grayed to indicate it is not implemented.

e) A "Last" button will reload the last dc/dt session, but this is not yet completely implemented.

8) In the gOfs report file, if weight averages were not checked on the dc/dt screen, a block of lines the same size will appear. If gOfs is checked, the calculated values will appear, regardless of their sign. BUG: some bad values may oflo the F8.3 Fortran output field.

 

9) The preference settings for ticks and grid are now being properly incorporated into new control files.

 

 

2/3/2009                           SedAnal v5.20

1) Scaling is shown in the grid winodw in the fitting screen if the "Show scaling details" box is checked. The max and min for each run can been seen, along with which points determined the scaling.

 

2) Some minor format changes were made in the general log output file.

 

3) Changes to the residuals window:

a) It no longer rescales the plot on each iteration if absolute scaling has been checked.

b) In the plot, radii which do not correspond to an experimental point are shown in a distinctive color, rather than being filled with the color corresponding to 0. In a fit with a single run, there will be at least one, and usually several, scans with no "fill" points; in a fit with multiple runs, there will be at least one run with a scan with no fill points. All scans for all runs are shown on the identical scale for radius, but the experimental radii fit depend on the user's choice of range to fit. This means that usually there is a difference between runs (and thus padding at one or both ends of the cell), and sometimes the individual scans in a run differ slightly in radii, which can also result in padding.

4) Changes to the script function:

a) Script selection loads the script and displays it; "Run script" starts it executing.

b) Cancel stops the script from running and closes the script control window. Using the script option on the main menu then works properly. NOTE: stopping the script does *not* stop the current fit, if one is in progress. If desired, that is stopped using the STOP button on the fitting screen, as usual.

c) The currently-active control file created by the script, if any, is displayed below the script.

d) A Help button shows some information on how to use a script.

e) The recent script files list now functions, to remember previously-run scripts.

 

 

1/26/2009                         SedAnal v5.19

1) Isodesmic fitting computes the correct weighted D, correcting an error introduced in v4.81.

 

1/25/2009        SedAnal v5.18

1) Running a script no longer overflows memory after several simulations that write .abr files.

 

1/20/2009                         SedAnal v5.17

1) You can now show ticks or a grid on the fitting screen.

a) There is a new button on the fitting screen toolbar to change the display while it's running.

b) There are new choices under Control extended/Fitting which select ticks, grid, or neither, for the x- and y-axes on the fitting plots. The default is in Preferences, and individual control files can be overridden in Advanced/Fitting.

c) Existing control files will apply the "Show ticks on plots" setting to both x- and y-ticks. Previously, this setting did nothing.

2) Plots on the fitting screen are show the run number only if there is more than 1 run.

 

3) The preprocessor now disables integral fringe subtraction for all but interference optics.

 

4) The fitting screen display has been changed to show the data set (original, subset N, or Monte Carlo N) being fit when L-M is used. Previously, this only showed when Simplex was used.

 

5) The appearance of the fitting screen toolbar buttons has been changed to improve their readability.

 

12/29/2008                       SedAnal v5.16

1) The convergence check in N-R has been fixed to force at least 1 iteration, rather than attempt to detect equilibrium from the initial guesses.

 

2) A new developer option has been added: Check equilibria. This option produces an output file <control name>_EquilCheck.txt, which, for each radial point and each sedimentation time step, verifies that mass is conserved, and that the calculated equilibrium constants match the actual ones (either fixed, user-specified, or floating parameters, current values). Only violations of these constraints are listed.

 

12/26/2008                       SedAnal v5.15

1) N-R returns the equilibrium concs, not the initial concs.

 

12/29/2008                       SedAnal v5.16

1) The convergence check in N-R has been fixed to force at least 1 iteration, rather than attempt to detect equilibrium from the initial guesses.

 

2) A new developer option has been added: Check equilibria. This option produces an output file <control name>_EquilCheck.txt, which, for each radial point and each sedimentation time step, verifies that mass is conserved, and that the calculated equilibrium constants match the actual ones (either fixed, user-specified, or floating parameters, current values). Only violations of these constraints are listed.

 

 

12/25/2008                       SedAnal v5.14

1) N-R is used when requested when only Keq's are given for the rxns. Previously, if there was no analytic sol'n to the equilibria, N-R was not used, regardless of the setting on the control Advanced/Fitting screen.

 

 

12/12/2008                       SedAnal v5.13

1) Deleted points are now handled properly. A hint on looking at deleted points: first, rescale by setting xMin, xMax and yMin, yMax to the region of interest (e g, 7.1,7.2 and 0.3,0.7). Then, if necessary, click on the region of interest to zoom.

 

2) The range to fit now uses only radii within the specified range (i e, r0 <= r <= r1). Previously, the points closest to the range boundaries were used, which could extend the range by a couple of points. This was only important when a range of points was deleted.

 

 

12/1/2008                         SedAnal v5.12

1) Another scrolling bug, which occurred when the base or meniscus was being fit, has been fixed.

 

2) The LM log file is now being closed at the end of a fit. This may possibly prevent the POSITION= error.

 

 

11/25/2008                       SedAnal v5.11

1) Master cells (first of a linked set) on the control screen can have their loading concs toggled between fit/hold when >10 runs have been scrolled. In previous versions, in some cases an error prevented the user from changing between fit and hold.

 

2) When scrolling > 10 cells on the control screen, incorrect, or blank, values sometimes appeared for some of the load concs. This has been fixed.

 

3) When a master cell had linked cells that were not visible, and the master cell's load concs were changed by typping new values, the values for those linked cells were not changed if they were not visible; when you scrolled down, the linked cells showed the previous master cell value. This error has been fixed.

 

 

 

11/19/2008                       SedAnal v5.09

1) The residuals window now works for equilibrium fitting. Previously it either game a constant value for all radii, or crashed.

 

2) Linking cells on the control screen now works properly. This error was introduced in v5.08, to correct another linking error--linking didn't work correctly for components > 4.

 

 11/19/2008                      SedAnal v5.10

This version is identical to 5.09, except I rebuilt it with the older Fortran compiler.

1) The residuals window now works for equilibrium fitting. Previously it either game a constant value for all radii, or crashed.

 

2) Linking cells on the control screen now works properly. This error was introduced in v5.08, to correct another linking error--linking didn't work correctly for components > 4.

 

 

 

11/13/2008                       SedAnal v5.08

1) Multi-speed data is now handled correctly in the preprocessor.

 

2) Equilibrium fitting and simulation are now correct (I hope!). Equilibrium simulation no longer gets an equation error.

 

3) Some errors that occurred on the control screen with a large number of both components and runs have been fixed.

 

4) The tab order on the bottom of the control screen has been redone.

 

5) Some information has been added to the general log. The general log is also useful for checking equilibrium fits, since it shows the molar loading concentrations used and the parameters dependent on them.

 

10/30/2008                       SedAnal v5.07

1) The display of curves during simulation has been fixed. In v5.03-5.06, these did not appear.

 

2) User data directories' names can now begin with a date mmddyy, and will be visible in the "Create new control file" window.

 

3) Preferences/General/CPU information, the field "OS support" now shows 0. The previous information was only applicable to x86 architecture (32-bit Windows).

 

4) The instruction windows that pop up in the Preprocessor when "Display instructions" is set in Preferences have been modified so they don't overlap or interfere with selection of radii.

 

5) In dc/dt

a) Switching from linear to log plot doesn't reload the data; selected scans are preserved.

b) Zooming and switching between linear and log now works, although depending on what part of the plot is zoomed, you may not see what you expect. Because the graphs are connected points, if a peak is truncated, there will be a line across its top.

c) When switching from Wide distribution to Normal, the right plot will be switched from log (for WDA) to linear.

6) When simulating, the file XLIinfo.IPS had the wrong values written into it for load concs other than the first component. This information was used only for the tooltip shown on the description of the .abr file when fitting using the simulated data. The problem has been fixed, but simulations done prior to 5.07 will be missing load concentrations for components 2,... .

 

7) The version identification, which appears in the title bar of the main menu and each function window, has changed.

Before: <version> <rewrite date>                    Example: v5.07 10/20/08

After: <version> <processor> <rewrite date> <O/S>   Example: v5.07 IA32 2008-10-20 XPPro

For the main menu option "Preferences", the window is now labeled "User preferences", and includes the above version identification.

 

8) An additional line appears below the title bar in the main menu. The title bar and this line contain the contact info, as well as the new version identification.

 

9) Incorrect scaling during simulation runs caused display problems, as well as "flat-line" plots, in v5.03-5.06. This has been fixed.

 

 

10/29/2008                       Model editor v1.76

1) The "calculate as" check boxes on the species pages no longer crash.

 

2) The legend has been repositioned slightly so it is completely readable.

 

 

 

8/2/2008                           Model editor v1.75

1) ME supports the two additional parameter types Ks2 and CM1.

 

2) On loading an existing model, the unused molecular parameters will appear as the defaults. Previously, these were blank. As before, the unused molecular parameters are grayed (disabled).

 

 

 

8/2/2008                           SedAnal v5.06

1) Two new molecular parameter types have been introduced: Ks2 and CM1.

 

2) SedAnal now computes non-ideal s and D according to the formulae

 

The developer option for choice of formula for s and D has been removed.

 

3) The Error estimation control now includes new initial increment boxes for fractional changes in F-stat params for Ks2 and CM1.

 

 

 

7/17/2008                         SedAnal v5.05

1) The plotting for multiple runs on the fitting screen has been corrected. The pagination (yellow arrows) also work, if the number of runs > number to plot on one screen.

 

2) New check boxes "Line" appear on the Preferences/Control extended/Fitting and Advanced/Fitting pages. These are not yet effective, but will be used to switch from points to lines on the fitting screen.

 

3) The controls on Advanced/Fitting pages which deal with line width and color, and point size, shape and color have been disabled, to indicate that the user cannot change these for an individual control file. They can only be changed via Preferences, when they will apply to all fits. The latter has always been true, but the user might have thought they were settable for each control file. See attached document for more information.

 

4) For the kinetic calculator, the header of the report has been modified to remove the inapplicable values. A sample of the new format is

  Report generated by SedAnal v5.05d  7/16/08         2008-07-16  17:11:48

  Run on P-NEW     This report file is [1]\kinetics m-di-tettramer_Report019.rtf

  Control file was [1]\kinetics m-di-tettramer.abc     Last modified 2008-07-16  17:10:50

| Kinetic calculator

|                                        Kinetic integrator errors:    0    0    0    0    0    0    0    0    0

|

|

|

 

Directories

[1] C:\BBRI\AUC\SedAnal\User_data\Kinetics\

The changed area is indicated by |.

 

5) In the preprocessor

a) the point size in the plot is controlled by the up-down arrows, as before, but there is a new button below it which toggles from points to lines.

 

 

b) In novice mode, typing in the "Meniscus at" box now works correctly. Before the caret vanished after each character.

 

 

7/11/2008                         SedAnal v5.04

1) A new choice kinetic integrator has been installed ("semi-implicit Euler extrapolation"), which seems to handle "stiff" ODEs, such as many kinetic cases, better. It may be selected under Advanced / Kinetics/equilibrium; choices are "BulSt", Bulirsch-Stoer and "SEulEx", semi-implicit Euler extrapolation. Semi-implicit Euler extrapolation is able to reach equilibria that Bulirsch-Stoer cannot, and is also much faster.

 

2) The kinetics and initial equilibration now operate properly for kinetic models with analytic solutions, and "use analytic soln" was chosen, that require >1 sec to reach equilibrium.

 

3) The "default is fit" wasn't being carried over from the ME to the control screen for molecular parameters other than ext coeff. This has been fixed.

 

4) The report for fitting did not show names of species >nRxns ( e g, 2 rxns and 4 species only listed 2 species); this has been fixed.

 

5) The behavior of the tooltip that shows when a screen shot has been written has been improved. Also, a tooltip appears when the mouse pointer touches the STOP button.

 

6) The tooltip showing s*, etc in the right window of dc/dt is now limited by the right edge of the plot. This tooltip still needs some work, as it flickers, and is hard to see at the top of the window.

 

 

 

6/4/2008                           SedAnal v5.03

1) The scaling for the radius is now done correctly for scans with significantly different numbers of scan points in the first and last scan pairs.

 

 

22-MAY-2008

SEDANAL v5.02

ModelEditor v 1.74

 

Numerous bug fixes and faster execution.

 

SEDANAL v 4.99

Preprocessor: option to plot lines instead of points added.

 

SEDANAL v 4.98

Preferences have been added to allow users to choose the method for computing mass action equilibrium concentrations for both the zero time point species concentrations (initial loading concentrations of all species) and the method to be used during the Lamm equation simulations.

 

DC/DT: Browsing for run files (*.abr) has been implemented.

 

Control screen: Time to reach the bottom ("bottom time") for the slowest species and "time-to-equilibrium" is shown for all models.

 

SEDANAL v 4.95

User scaling of in the residuals bitmap window.

 

SEDANAL v 4.92

Kinetic calculator now shows calculated equilibrium constants as a check.

 

SEDANAL v 4.91

Preprocessor will now browse correctly for scan files.

 

New control variables for calculation of equilibration values of initial species concentrations for sedimentation velocity of interacting systems.

 

SEDANAL v 4.89

Density now taken from the experimental data file (see preprocessor). (default=1.000)

"Viscosity" changed to "Viscosity ratio".

Output log includes additional concentrations.

 

2-Feb-2008

SEDANAL v4.88

ModelEditor v1.72

 

The ME now allows you to specify molecular parameters consisting of any three from (M, s, densIncr and D) for sedimentation velocity models.

 

17-JAN-2008

SEDANAL v4.86

ModelEditor v1.70

 

User auxiliary symbols are available. You must have installed Model editor v1.70 or later for this.

 

The residuals for the bitmap display are now interpolated onto a uniform grid whose size is the size of the plot window, and whose range is the range to fit selected in the preprocessor. The residual window can be resized by dragging its edges, and the size of the residual plot changes to exactly fit the window horizontally.

 

Doing a simulation can create a prototype control file for fitting. The name of this file is <simulation control filename>_FIT.abc. For example, if the control file for simulation was MonomerDimer.abc, the prototype control file for fitting will be MonomerDimer_FIT.abc. This is settable in the preferences.

 

Preprocessor plotting: You can now change the point size on the plot. The up/down control next to the Plot button, allows you to choose between filled or hollow circles, solid line and points. The default is points (as it was in previous versions).

 

Model Editor:

    The maximum length of molecular parameter names is slightly shorter. There is a new field for each molecular parameter, "Symbol". This refers to the symbol used in the equation editor (control screen). It is mainly useful for user auxiliary symbols.

 

24-NOV-2007

SEDANAL v4.85

ModelEditor v 1.69

 

Now dc/dt WDA always includes an extra table at the end, showing all the curves interpolated onto the same uniform grid and averaged. (This will be settable in the preferences in the future)

 

29-OCT-2007

SEDANAL 4.84

 

Error handling of bad-format scan files in the preprocessor has been improved.

 

Simulation can now generate a cell data (run, or .abr) file, as well as the usual scan files. The run file's name is YYYYMMDD_simHHMMSS.abr, and it has the meniscus and cell base from the simulation, and the range to fit is (meniscus+0.05) - (base-0.1) (i e, 5.95-7.1 for m=5.9 and b=7.2).

 

 

13-OCT-2007

SEDANAL v4.82

 

A "residuals" bitmap window is available on the fitting screen. It can be toggled on/off by clicking the "R" button at upper right.

 

3-SEP-2007

SEDANAL v4.81

 

Loading concentrations can now be entered in either mol/L or g/L.

 

For sedimentation velocity fitting, any 3 of the 4 parameters s, D, M, density increment can be specified (in the model); the missing one will be computed using the Svedberg equation at the start of each iteration.

 

14-June-2007

SEDANAL v4.80

 

preprocessor now handles multiple-wavelength data and

writes separate run (*.abr) files for each wavelength.

 

5-APR-2007

SEDANAL v4.73

 

Change in the non-ideality of s and D:

s(i)=so(i)/nonIdeal(i)

There is a choice between (1) and (2):

nonIdealKs(i)=1+Ks(i)*cTotal       (1)

nonIdealKs(i)=1+SUM(Ks(i)*c(i))    (2)

and

D(i)=Do(i)/nonIdealKs(i) * nonIdealBM1(i)

nonIdealBM1(i)=1+2*BM1(i)*cTotal      (3)

nonIdealBM1(i)=1+2*SUM(BM1(i)*c(i))   (4)

For D, there is a choice of (1) and (3), or (2) and (4).

 

25-Feb-2007

SEDANAL v4.6x

ModelEditor v1.69

 

Scripting ability added

Kinetics simulator added.

WDA screen widened.

 

15-May-2006

SEDANAL v4.37

ModelEditor v1.67

 

General improvements and improved F-statistics search algorithm.

 

1-March-2006

SEDANAL v4.33

ModelEditor v1.67

 

Some remaining bugs have been fixed.

 

10-Jan-2006

SEDANAL v4.25

ModelEditor v1.67

 

            Levenberg-Marquardt minimization routine added to speed up fitting.

 

            An Equation Editor was added to the Control Screen to allow arbitrary constraints relating global and local parameters.

 

9-June-2005

SEDANAL v4.10

 

            Determination of confidence limits using F-statistics has been implemented.

 

 

15-May-2005

SEDANAL v 4.02

ModelEditor v 1.65

 

BIOSPIN, the sedimentation equilibrium analysis program of Roark and Yphantis (1969) has been implemented as a subroutine of SEDANAL.

 

The preferences have been expanded to allow setting of commonly used default parameters for various SEDANAL functions.

 

15-June-2004

SEDANAL v 3.61

ModelEditor v 1.61

 

            Global fitting to general self- and hetero-associating systems has been implemented.

 

25-March-2004

SEDANAL v 3.45

ModelEditor v 1.59

 

Fitting to sedimentation equilibrium data from non-interacting systems composed of up to 4 components has been implemented. The preprocessor will read run files and blank files and perform the appropriate interpolated optical blank subtraction as a function of radius. Analysis of any interacting system that can be represented in the ModelEditor will be coming. At present, this feature is undocumented.

 

A new Main Menu item has been added called “Equilibrium Calculations” which allows one to compute the equilibrium concentrations of all species at any given total concentration of components and values for the equilibrium constants. It is independent of any sedimentation and represents the initial distribution of species at the beginning of the run.

 

On the main fitting screen, (i.e. the control screen) there is new button called “Output” that allows in this version the writing of a file with the kinetics time course to equilibrium for the model loaded into SEDANAL from any abc file. It prints out a file in columns of the time and the concentration of each species.

 

The ModelEditor now uses the density increment [i.e. or (1-vr)] by default instead of partial specific volume. Either parameter can be used. Just remember that since the solvent density is set to 1.0000 internally that partial specific volume is actually partial specific volume multiplied times the solvent density, as before.

 

An isodesmic model has been added to the ModelEditor. Provisions have been made in SEDANAL to allow one to enter coefficients for a polynomial representing the functional dependence of sedimentation coefficient upon the degree of oligomerization.

 

An isoenthalpic model has also been added. Use it with caution: it has not been tested thoroughly yet.

 

The output file for the Wide Distribution Analysis has been changed to include column headers.

 

The maximum number of components has been increased from 3 to 4. The maximum number of species is still 7 and the maximum number of reactions is still 6.

 

Both Monte Carlo and Bootstrap with replacement have been added to the “advanced” menu on the Control Screen.

 

A provision has been added to the control screen of SEDANAL to compute and output the time course of the kinetics of any reaction schemes that can be represented in the ModelEditor.

 

Improvements have been made to the Report file.

 

There have been numerous other small enhancements and bug fixes.

 

The fitting screen may be printed at any time by right-clicking on the STOP/CLOSE button. The screen may be written to a bit map file (with same name as the report file) by right-shift-clicking the STOP/CLOSE button.

 

15-OCT-2003

SEDANAL v3.15

 

            A boot-strap with replacement algorithm has been added to compute standard deviations of the fitted parameters automatically.

 

            On the Control Parameter screen under “Advanced…”, one can set a ”stop fitting” criterion and enter the number of Boot-strap cycles to perform.

 

            An extensive report is now generated at the end of a fitting session. The report includes a log and summary of the boot-strap results.

 

            A number of small bugs and “irritating features” have been remedied.

 

 

07/22/2003

 

Changes in v2.97

 

Preferences can be changed from the Main menu.

 

A bug in the monomer-dimer analytical mass action solution has been fixed

 

Default choice of scans for DCDT has been set to a small group of scans near the middle of the data set. This results in a faster initial plot.

 

Major changes in the preprocessor have been made to accommodate Wide Distribution  Analysis.

 

06/10/2003

 

v2.99

 

Time derivative multi-speed wide-distribution analysis (WDA) is now incorporated into SEDANAL.

 

WDA is carried out essentially by observing concentration changes at a particular radius (or set of radii) and converting the time rate of change of the local concentration into a plot of g(s*) vs s* at those radial positions. The radial positions are chosen as a compromise between resolution (higher radius) and the observing the slowest sedimenting material (lower radius).

 

A bug in the monomer-dimer model has been fixed.

 

 

06/10/2003

 

Changes in v2.91

 

Time derivative g(s*) analysis (DCDT) has been added to SEDANAL.

 

There is a new version of the ModelEditor (v1.4)

 

02/18/2003

 

Changes in v2.79:

 

A large speed improvement (as much as 20-fold for some models) has been achieved by a combination of the removal of some debugging code and the inclusion of analytical solutions to the mass action calculations for several common models.

 

Analytical solutions for both kinetics and equilibrium are available for the models A=A2, A=B and A+B=C.

 

SEDANAL will recognize these models and will use the appropriate analytical expression: it will use the equilibrium analytical expression if an equilibrium constant is given, and the kinetic analytical expression if the rate constants are specified on the control screen. Models which do not have analytic solutions use the ODE (Ordinary Differential Equation) integrator, as before.

 

2) There is a new version, 1.2, of the ModelEditor that has some bug fixes regarding recognition of components for some complex reaction schemes.

 

 

01/10/2003

 

SEDANAL v2.72

 

            In addition to the ability to fit to data from multiple optical systems introduced in v2.70, the ability to constrain the loading concentrations to be the same for all data sets from the same cell has been added. See below for details.

 

Also the plotting problem with WindowsXP was tracked to the OS and was fixed with a temporary kluge. It may reappear in later versions on faster machines.

 

12/20/02

 

SEDANAL v2.70

 

            The ability to fit globally to data from multiple optical systems has been added to this version. The box for the extinction coefficient on the Control Screen has been changed to allow a different extinction coefficient for each cell (or run).

 

For extinction coefficient, there is a single row of values (each under the column for each of the species). The initial setting is for "ALL CELLS". To change the row being shown, left-click on the parameter label ("Mass extinction coefficient/ALL CELLS").

 

Changing values in the "ALL CELLS" row makes the identical change in all other rows, so the correct procedure is

 (1) Determine the predominant optical system.

 (2) Enter the extinction coefficient for this system in the "ALL CELLS" row.

 (3) Enter the different extinction coefficient for the others in the other cells' rows.

Note that if an error in step (1) is discovered later, all coefficients must be entered from the beginning, since changing the "ALL CELLS" row overwrites all the individual cells' rows.

 

11/22/02

ModelEditor 1.1

 

Updated to manage components better.

 

SEDANAL v2.66

 

            This version has been accelerated considerably by eliminating much of the screen I/O.

 

1) If there is an obvious error in the choice of components user will get the message

 

“Fitting cannot proceed due to errors - see log file for details”

on the screen, and

“*** Terminating fitting because n model component(s) are incorrect”

in the log file.

 

2) A new configuration file, Preferences.txt in the SEDANAL \Program\ directory, contains the paths to the User_data directory and the model editor. Soon to come will be the ability to edit this file.

 

3) The Preferences.txt file contains the maximum number of scans to plot on the initial input of centrifuge data (30 in the preferences file included). If there are more scans, it plots 30 approximately-equally-spaced scans. This becomes the default for future pre-processing, unless one changes it, and is stored in the run file.

 

4) Selected scans may be fitted. The default is all scans; to change this, right-click on the run file name (lower left) and select the scans. The result is written into the control file. Old control files default to "all scans". There is an option in the scan-select dialog box to use the scans chosen for plotting for fitting also.

 

11/18/02, v2.64

 

1) The equilibrium calculation is now done correctly in the case of components which are not the lowest-numbered species.

 

2) Ratios are now molar, based on the MWs at the time of the iteration (if MWs are being fitted, they will vary). A later version will allow choice of concentration units for the display. Internally all concentrations are on the mass scale.

 

3) The display is now updated once every 10 seconds, when a new minimum is encountered, or when one forces redisplay by minimizing and maximizing, etc.

 

11/14/02, v2.63

This version is the same as v2.62 except that .RAS simulated output files have an 'R' instead of 'P' on the header line.

 

11/11/02, v2.62

 

1) Fitting control screen display of floating point numbers has been changed to eliminate leading 0s in the exponent; i e, what was 1.23456e-0005 (which got truncated in the box) is now 1.23456e-5.

 

2) Dimensioning has been corrected to be uniformly 1600 points maximum in the grid.

 

3) Typing RETURN when in one of the 4 scaling boxes on the preprocessor screen is the same as pressing PLOT (i e, starts re-plotting).

 

 

2002/10/16.

 

 

ModelEditor 0.8

 

            Has a “copy” function for parameters that will be the same for all species. In the previous versions, derived parameters, like partial specific volume and molar mass were calculated for other species either as a sum or weight average of other species. The “copy” function allows the same value of a parameter to be used for several species.

 

 

SEDANAL v2.59

 

            Minor bug fix. The log files were sometimes being written into the Program folder. This has been fixed so that all output files are written into the “experiment” folder now.

 

 

SEDANAL v2.58

 

We’ve added the ability to fit for the ratio of the loading concentration of one component relative to that of another as a global parameter over all cells and is used in cases in which the runs are part of a dilution series for which that ratio is usually expected to be constant.

 

We’ve added the ability the fit the ratios of equilibrium constants. For example, in the case of two-step binding where the ratio of binding constants could either be allowed to float or held constant. For putative, two-step, non-cooperative binding (the null hypothesis in most cases) the ratio of K₂/K₁ would be 0.25. It would be convenient to specify that ratio while letting the value of K₁, and, therefore, K₂, vary accordingly.

 

SEDANAL now allows fitting to systems composed of 3 components. This feature allows fitting for molar mass in a mixture of up to three non-interacting species. Also, allows inclusion of an aggregate or other additional, non-interacting component in a two component interacting system like A+B=C.