User`s Manual - MikeBlaber.org
Contents
1. IV TD Formally it is the temperature at which the native and denatured states are equally populated TD must exist In other words the model requires that there be some definite temperature at which the native and denatured states are equally populated If the conditions are such that the native is never more than 50 populated the analysis will fail V k The k parameter is a scalar term applied to the raw data molar heat capacity If it is allowed to float during refinement and does not refine to a value of 1 0 there are two possible interpretations 1 AHw AHea 1 0 but the concentration applied during normalization of the molar heat capacity is incorrect In this case the actual concentration of the protein in the DSC analysis is the value used during the normalization k Furthermore the refined thermodynamic parameters are for the corrected concentration 2 The concentration is correct but AHw is not equal to AHea In this case the value of k is equal to AHyy AH ai Furthermore the refined value of DDTD is equal to AHyy AHca can be derived by dividing DDTD van t Hoff enthalpy by k Thus k should only be allowed to float if other data Is available to confirm either the concentration or the value of AHy AH a See Advanced Topics section DSCFIT v 15 3 User s Manual 03 25 2002 http wine1 sb fsu edu DSCfTit Section 3 Available Models DSCFIT is customizable to allow a variety of treatmen2. First order term for Native State Bol 436 449870347 Zero Order term for Native State Co 128027 72382 Second order term for Denatured State 41 fo First order term for Denatured State E1 3 969081 4087 Delta Cp at Melting Temperature OCTO Jeagr 17 704232 Enthalpy at Melting Temperature DOTE 257266 416971 Melting Temperature TO a 2 566r03905 Concentration Constant K f Print Results Now A is allowed to float during refinement thus Cp T will be a second order polynomial 18 DSCFIT v 15 3 User s Manual 03 25 2002 http wine1 sb fsu edu DSCfTit Set Parameters Reman Fised T Second order term tor Native State Ac il First order term for Native State Bo 417456915434 Zero Order term for Native State Co 128673 63922 Second order term for Denatured State A1 0 7301307263 First order term for Denatured State E1 128 1 74880682 Delta Cp at Melting Temperature DETO Jes7e 44296027 Enthaloy at Melting Temperature POTE 2568322627223 Melting Temperature TD 312 435522846 Concentration Constant K Print Resuts a O ee xl The fitted function and Cy T and Cp T functions look like this note the curve of the Cp T function fe DSCFit DSCFitt File Edit View Functions Window Help x E Load Data Graphs Refine Evaluate Parameters Export Heat Capacity Graphs 19 DSCFIT v 15 3 User s Manual 03 25 2002 http wi
3. Load Data menu command or the Load Data command on the menu bar A dialog window will open enabling you to specify the desired data file the default file extension is dat When reading in the data file DSCFIT will present you with a dialog box to allow conversion of C or calorie based data to K and Joules Units DSCFit only uses units of Kelvin and Joules Ifthe data loaded does not have appropriate units they must be converted My data s units are Temperature Units Energy Units amp Kelvin amp Joules C Celsius C Calories LL a Select the appropriate radio buttons for the units of the input data in cases where the temperature appears to be in C rather than K DSCFIT will alert the user to the need to convert The temperature and heat capacity units are independent Thus if for some bizarre reason your raw data is in C and Joules select the radio buttons for Celsius and Joules DSCFIT v 15 3 User s Manual 03 25 2002 http wine1 sb fsu edu DSCfit After reading the data file DSCFIT will report the number of successfully imported data points and the temperature range of the data 214 data points were successfully loaded The data s temperature ranges from 2o4 Kto adz K Before continuing be sure that this information is as expected for the data file in question Note Currently DSCFIT applies the conversions of K 273 15 C and Jou
4. User s Manual 03 25 2002 http wine1 sb fsu edu DSCfTit Constrain Data C All Data Points Constraints amp Constrain Points W Mo Points less than W No Points greater than Select the Constrain Points radio button to constrain the region of data for analysis then enter the low and high temperature cutoff values in K When the data is subsequently graphed the cutoff data regions will be displayed in green and the active data points remain blue oie DSCFit DSCFitt File Edit View Functons Window Help _ o x Load Data Graphs Refine Evaluate Parameters Export Heat Capacity Graphs d 102000 113400 1244200 135000 i 290 300 310 320 330 340 a Il Evaluation of Non 2 state Behavior Various criteria are used to determine whether the experimental data agrees with the 2 state assumption of the model The first relates to the deviation of the raw data from the refined model This deviation is evaluated by 13 DSCFIT v 15 3 User s Manual 03 25 2002 http wine1 sb fsu edu DSCfTit noting the value of the Standard Deviation in the Refine Dialog Box and comparing this value to the expected instrumentation noise Convergence Criteria Convergence Criteria Number oflterations MO C Iterations with no Change i I C Delta Chi oon Standard Deviation 50 Error Analysis Chi E1891 10 6 Last Delta Chi 2 2 1 364e 005 Standard Deviation fi rE 534 7425864 Ite
5. for Native State Co 123418 92982 z Second order term for Denatured State A1 oo E First order term tor Denatured State B1 174 891326096 E Delta Cp at Melting Temperature CETO 1006 5456466 E Enthalpy at Melting Temperature POTO 262950118815 E Melting Temperature TO 214 070507679 Print Results In this case residual structure leads to residual enthalpy after the main transition This causes an aberrantly large positive slope for the Cp T function Thus at the TD the Cp T function may actually appear to be lower than the Cy T function resulting in a negative value for ACp Finally the k parameter can be refined instead of being constrained to 1 to evaluate the possibility of non 2 state behavior In this case it is essential that the concentration of the sample be accurate f k is allowed to float during refinement and does not refine to a value of 1 0 there are two possible interpretations 1 AHw AHaea 1 0 but the concentration applied during normalization of the molar heat capacity is incorrect In this case the actual concentration of the protein in the DSC analysis is the value used during the normalization k Furthermore the refined thermodynamic parameters are for the corrected concentration 2 The concentration is correct but AHw is not equal to AHea In this case the value of k is equal to AHyy AH ai Furthermore the refined value of DDTD is equal to AHyy AH lt a can be derive
6. standard deviation of the fit decreases to a specified value Additionally there is a button labeled Coefficients Selecting this button will bring up the Coefficients Dialog Box where the values of the various thermodynamic parameters can be manually entered or the constrain check boxes may be selected or deselected Clicking the button labeled Begin will start the refinement of the thermodynamic parameters to improve the fit with the raw data The refinement will proceed until the selected criteria for convergence has been met If more than 100 cycles of refinement are performed without the convergence criteria being met the program will exit from refinement with an alert message DSCFIT v 15 3 User s Manual 03 25 2002 http wine1 sb fsu edu DSCfTit Statistical information on the refinement is available within Error Analysis section of the Refine Dialog Box Convergence Criteria Convergence Criteria iterations with no Change i U C Delta Chi 0 00 C Standard Deviation 50 l es Chi E1 a91 7 1076 Last Delta Chi 2 1 364e 005 Standard Deviation f rE 534 7425864 Iterations since last change f 0 Coefficients E mit The information includes 1 the value of the yf merit function 2 the last value observed for the change in the merit function 3 the standard deviation of the fit to the experimental data and 4 the number of iterations that have occurred since an improvement in the merit
7. the fitted function and the raw data It is useful for determining the magnitude of the error i e whether it is within the expected instrumentation noise level and whether the error is random or systematic indicating possible non 2 state behavior DSCFIT v 15 3 User s Manual 03 25 2002 http wine1 sb fsu edu DSCfit Delta C Delta H Delta G Delta S plots for these thermodynamic functions On the left hand side of the Graphs Dialog Box are parameters used in graphic the various functions The default values are to autoscale Y and the temperature range is the minimum and maximum of the data points V Refine Dialog Box After the appropriate graph has been selected the graph will be shown along with the Refine Dialog Box Convergence Criteria Convergence Criteria Iterations with no Change fic C Delta Chi pom o C Standard Deviation o0 Eror napsi Chie B 8287400 7 Last Delta Chi 2 oo000 Standard Deviation 545 0603071194 lterations since last change fo Coefficients E mit The Refine Dialog Box allows you to select the criteria for convergence of the model fit to the experimental data The choices include 1 refine for a fixed number of cycles or 2 refine until no further improvement in the merit function is observed over a given number of refinement cycles or 3 refine until the improvement in the merit function decreases to a specified value or 4 refine until the
8. 223306 What happens to the fit if we include the same scalar k as before Here are the results of the fit 26 DSCFIT v 15 3 User s Manual 03 25 2002 http wine1 sb fsu edu DSCtTit simulated DSC Data H_ H_ 2 0 18 DSC Fit with Concentration Parameter a Oo a a Parameter Ao 0 Ay 0 Bo 0 B 0 Co 0 k 1 99901503 DCTD 0 DDTD 334715 0703 TD 323 1500011 In this case DDTD agrees with the expected value for the van t Hoff enthalpy 334 9 kJ mol The value for the parameter k agrees with the ratio of AHyH AH ai 2 0 AHcai Can now be derived by dividing DDTD van t Hoff enthalpy by k AHyH AH cai yielding 167 4 kJ mol Conclusions e Errors in concentration or situations where the van t Hoff enthalpy do not agree with the calorimetric enthalpy can result in a similar situation errors in enthalpy and deviation of the fit with a 2 state model e Introduction of a scalar k in the model can correctly adjust the model for two situations 1 errors in concentration assuming van t Hoff and calorimetric 27 DSCFIT v 15 3 User s Manual 03 25 2002 http wine1 sb fsu edu DSCtTit enthalpies are equal and 2 determination of the van t Hoff to calorimetric enthalpy ratio assuming concentration is correct section 6 Output of Function Data for Use in Other Programs Data associated with relevant thermodynamic functions can be output for use in other programs This is particularly useful
9. 3 25 2002 http wine1 sb fsu edu DSCfTit References Kidokoro S l and A Wada 1987 Determination of thermodynamic functions from scanning calorimetry data Biopolymers 26 213 229 Freire E and Biltonen R L 1978 Statistical mechanical deconvolution of thermal transitions in macromolecules Theory and application to homogeneous systems Biopolymers 17 463 479 Referencing DSCFit in your publications Details of the DSCFit program have been published in the following An Efficient Flexible Model Program for the Analysis of Differential Scanning Calorimetry Data Grek S Davis J and Blaber M Protein and Peptide Letters 6 429 436 2001 29
10. 5983 B 152 5561894 152 5561889 152 5561915 Co 58243 45094 58243 45113 58243 45016 k 1 526433462 0 7632167336 0 3816083605 DCTD 2273 729531 2273 729583 2213 729321 DDTD 271551 4875 271551 4878 271551 4863 TD 316 0675458 316 0675458 316 0675458 The parameters refine to identical values except for the k parameter which reflects the concentration error in the normalization Why is the value for k not equal to 1 0 for the data set normalized to 0 04mM One possibility is that our sample is not actually at 0 04mM but is actually 0 052mM Alternatively the protein may not exhibit two state denaturation and thus the van t Hoff and calorimetric enthalpies may not be equal i e AHw AHeca does not equal 1 0 How does a non equality of the van t Hoff and calorimetric enthalpies affect the analysis Here is a fit to a simulated DSC run with the following parameters 25 DSCFIT v 15 3 User s Manual 03 25 2002 http wine1 sb fsu edu DSCtTit TD 323 15 K Hea 167 4 kJ mol Hyy 334 9 kJ mol AHyH AHea 2 0 ACp 0 slopes for baselines 0 Simulated DSC Data H_ H__ 2 0 ue DSC Fit kJ mol K 2490 300 210 220 230 340 300 360 Temp K The fit undershoots the peak indicating a situation where the AHw Hea iS greater than 1 0 with possible multimers in the native state In any case a two state model does not fit the data The parameters of the fit are Parameter OOOO 0 CO DDTD 220899 7025 TD 323
11. DSCFIT Flexible Rapid and Automated Analysis of Differential Scanning Calorimetry Data USER S MANUAL Version 15 3 3 25 2002 Sasha B Grek John K Davis and Dr Michael Blaber Institute of Molecular Biophysics Florida State University Tallahassee FL 32306 4380 TEL 850 644 5870 FAX 850 561 1406 This work was supported by a grant from the Florida Space Grant Consortium htto wine1 sb fsu edu DSCrfit DSCFIT v 15 3 User s Manual 03 25 2002 http wine1 sb fsu edu DSCtTit Table of Contents Program Description 2 System Requirements and Limitations 2 How to Use this Manual 2 section 1 General Steps in Data Analysis Loading of Data 3 ll Initial Guess of Thermodynamic Parameters 4 Ill The Graphs Dialog Box 9 IV The Refine Dialog Box 6 Section 2 Description of Thermodynamic Parameters I Ao Bo Co 8 ll A1 B1 DC 8 Il DD 9 IV TD 9 V K 9 Section 3 Available Models l Cy T 10 Il Cp T 10 Ill DCTD 10 IV DDTD and TD 11 V k 11 Section 4 Advanced Topics I Data Clipping 11 II Evaluation of Non 2 state Behavior 13 III Use of Second order Polynomial Functions for Cn T or Co T 18 IV Determination of Thermodynamic Constants at Temperatures other than TD 20 V More than One Possible Solution for Refined Parameters 20 VI Situations Where TD Does Not Exist 21 Section 5 Derivations 22 Section 6 Output of Function Data for Use in Other Programs 28 References 29 Re
12. ay a Note this equation should model your raw heat capacity data A Modification to Allow for the Determination of the van t Hoff to Calorimetric Enthalpy or to Determine the Error in Concentration of the Protein Sample What happens if an inaccurate value for the protein concentration is used during the normalization of DSC data How does this affect the fitted thermodynamic parameters and how can this be evaluated in the analysis In the following example the protein in the DSC experiment is determined to be 0 04mM However during analysis of the data the heat capacity is normalized to three different concentrations 0 04mM correct 0 08mM and 0 02mM The resulting endotherms look like this mo 100 110 120 130 140 150 160 170 180 190 200 kJ mole deg Normalized to 0 08mM Normalized to 0 04mM Normalized to 0 02mM 290 300 310 320 330 340 390 Temperature K Fitting our model to this data we get the following results 23 DSCFIT v 15 3 User s Manual 03 25 2002 http wine 1 sb fsu edu DSCtTit 0 04mM Protein Sample Normalized to either 0 02 0 04 or 0 08mM Normalized to 0 08mM Normalized to 0 04mM Normalized to 0 02mM 290 300 310 320 330 340 350 Temp K The fit overshoots the data normalized to 0 02mM suggestive of a folding intermediate fits reasonably well with the data normalized to 0 04mM the correct concentration and undershoots the data normal
13. d by dividing DDTD van t Hoff enthalpy by k Parameter refinement of the prior data set with k allowed to float results in a value for k of 0 80 indicating a AHyH AHca ratio consistent with non 2 state behavior i e the presence of an intermediate state DSCFIT v 15 3 User s Manual 03 25 2002 http wine1 sb fsu edu DSCfit Thus the evaluation of non 2 state behavior relies upon analysis of e The magnitude of the standard deviation e The nature of the residual scatter random vs systematic e The value of ACp at the TD e The value of the refined K parameter i e AHyy AHga1 Ill Use of Second order Polynomial Functions for Cy T or Cp T DSCFIT allows the option of the use of second order polynomial functions for the Cx T and Cp T functions The general paradigm of protein denaturation for example is that the Cy T function is well approximated by a linear function while the Cp T function may exhibit slight curvature However over the typical temperature ranges used to study protein thermal denaturation this curvature is slight and a linear function is a reasonable approximation for the Cp T function The following is an example of the application of a second order polynomial for the Cp T function for a data set of human acidic fibroblast growth factor First the refinement is performed with a linear function for both Cy T and Cp T Set Parameters Remain Fired Second order term for Native State Ao fo
14. erature TD a 2 566r 03903 Concentration Constant K f Print Results E me m e a m m x Section 2 Description of Thermodynamic Parameters Ao Bo Co The native state heat capacity function Cn T is defined by the following function Cy T 3 Ag T TD 2 Bo T TD Co where TD is the melting temperature Thus Cy T is a second order polynomial when Ao is non zero If Ao is constrained to a value of 0 default value for Ao Cy T is a linear function Likewise if both Ag and Bo are constrained to values of 0 Cy T is a constant In any case Co is the value of Cy T at the TD ll A41 B4 DOTD The denatured state heat capacity function Cp T is defined by the following function Cp T 3 Ay T TD 2 B T TD DCTD Co DSCFIT v 15 3 User s Manual 03 25 2002 http wine1 sb fsu edu DSCfTit where TD is the melting temperature Thus Cp T is a second order polynomial when A is non zero If A is constrained to a value of 0 default value for A1 CD T is a linear function Likewise if both A and B are constrained to values of 0 Cp T is a constant DCTD is the value of ACp T at the TD Since ACp T is defined at Cp T Cn T and Co is the value of Cy T at the TD DCTD Co is equal to the value of Cp T at the TD Hl DDTD DDTD is defined by the following function AH T A1 Ao T TD B1 Bo T TD DCTD T TD DDTD Thus DDTD is the value of AH T at the TD
15. ferencing DSCFit in your publications 29 DSCFIT v 15 3 User s Manual 03 25 2002 http wine1 sb fsu edu DSCfTit Program Description DSCFIT is designed to analyze data from differential scanning calorimetry DSC experiments It uses a statistical mechanics based two state model in combination with an optimized non linear least squares fitting routine The program is designed to be flexible rapid and require a minimum of operator intervention By setting appropriate parameters users can customize the type of model to be used in fitting the data DSCFIT has the unique ability to evaluate concentration errors in DSC data in addition to the determination of van t Hoff and calorimetric enthalpies system Requirements and Limitations DSCFIT is a Windows 95 98 NT based program A graphics display of at least 800x600 is required A minimum of 32 MB of system memory is also required The disk space requirements of the program are quite modest currently lt 5 MB As currently implemented DSCFIT can analyze data sets with 2000 or fewer data points Data from both upscan and downscan experiments can be analyzed Please report all bugs etc to Dr Michael Blaber Institute of Molecular Biophysics Florida State University Tallahassee FL 32306 4380 TEL 850 644 5870 FAX 850 561 1406 blaber sb fsu edu Suggestions for improvements to DSCFIT are also welcome How to Use this Manual If you are ready to analyze a data set pr
16. for presentation quality graphics To output function data choose the Functions Export File command A DSCFit DSCFit1 File Edit View Bitte Window Help Graphs phs Refine Evaluate Parameters Export Refine Evaluate Point Heat Capacity Grapns Parameters 102600 Constrain Data Export File h 113400 You will be given the option of adding comments to the top of the output file You can output the file as either a text or csv file which can be read directly into popular spreadsheet programs The output of a typical csv file will look like this Ao Bo Co A1 B1 DCTD DDTD TD k 0 99 9690 121530 0 435 449 6497 17 257266 312 5667 1 Temperature Raw Data C T Residual CpN T CpD T DCp T DH T DS T DG T Fn T Scatter 287 276 149246 148519 8 726 19735 126587 150053 23466 121631 451 06386 7947 93 0 034633 287 55 148989 148426 5 562 50016 126532 149814 23282 115227 428 78033 8068 47 0 033087 287 823 148713 148316 5 396 49734 126477 149577 23099 108896 406 77338 8182 52 0 031694 288 097 148446 148191 07 254 92639 126422 149338 22915 102592 384 8814 8290 97 0 03043 288 373 148220 148051 33 168 67178 126367 149098 22730 96293 363 02736 8394 17 0 029279 This output data includes the input raw data fitted Cp T function residual scatter native and denatured state baselines AC T AH T AS T AG T and Fy T fraction native state functions 28 DSCFIT v 15 3 User s Manual 0
17. function was observed Each time the refinement routine identifies values that improve the agreement of the model with the raw data the currently displayed graph will be updated When is the refinement of the thermodynamic parameters complete As refinement proceeds the value for the y merit function should decrease although a target value is difficult to know in advance The value for Ay should also decrease with values typically smaller than 1 x 10 typical of convergence The standard deviation should also decrease Target values for the standard deviation would be related to the expected error for the instrument e g on the order of 100 200 J mol K If the fit has converged many cycles i e gt 10 of refinement will occur without an improvement in the fit Clicking on the Coefficients button in the Refine Dialog Box will bring up the Coefficients Dialog Box from which the values for the refined thermodynamic parameters can be read DSCFIT v 15 3 User s Manual 03 25 2002 http wine1 sb fsu edu DSCfTit Set Parameters Remain Fired x Second order tern for Native State o fo First order term for Native State Ba 435 449870637 ero Order term for Native State Co 128027 72383 Second order term for Denatured State 41 fo First order term for Denatured State E1 s9 9690809707 Delta Cp at Melting Temperature OCTO 6497 17706293 Enthalpy at Melting Temperature DOTE 257266 416056 Melting Temp
18. is reference temperature will be displayed V More than One Possible Solution for Refined Parameters Two cases have been observed where alternative solutions have been obtained during refinement Both cases can occur when the initial guess values are poor This situation will typically not occur if the default initial guess values are used but may occur with operator selected values The first case is a mathematically equivalent solution but where the sign of DDTD and DCTD parameters are inverted This can be demonstrated using the data set from the 20 DSCFIT v 15 3 User s Manual 03 25 2002 http wine1 sb fsu edu DSCfTit discussion of the previous section inverting the sign of DDTD and DCTD and refining Set Parameters Remain Fised wf Second order tern for Native State Ao fo E First order term for Native State Bo 39 9690820457 E ero Order term for Matie State Co 121530 54685 m Second order tern for Denatured State A1 fo E First order term for Denatured State B1 435 449869195 E Delta Cp at Melting Temperature CETO 6497 1769654 E Enthalpy at Melting Temperature POTO 257266 41741 E Melting Temperature TO 212 566703927 Print Results Note that an equivalent solution has been found for the magnitude of DDTD and DCTD parameters but that the signs are inverted The fitted function will be equivalent to the correct solution The second situation involves an apparent loca
19. ized to 0 08mM suggestive of a van t Hoff enthalpy greater than the calorimetric indicative of multimers in the native state The refined parameters for the above fits are as follows Parameter Normalized to Normalized to Normalized to 0 08mM 0 04mM 0 02mM Ao 0 0 0 Ay 0 0 0 Bo 106 399543 491 9808408 1310 731465 B 142 2004446 146 4605106 12 70224628 Co 42033 46479 12659 20659 133991 364 DCTD 3276 45386 1771 328293 3131 276108 DDTD 235633 051 297758 6783 385858 965 TD 315 5935796 316 2024851 316 1545567 If a scalar is introduced into the basic expression for the heat capacity function the error in the normalization of the concentration can be corrected aa a 2a E a e a Ie k RT 24 DSCFIT v 15 3 User s Manual 03 25 2002 http wine1 sb fsu edu DSCfit Here we are dividing our model for the experimentally observed heat capacity by the scalar term k If the above three data sets are re analyzed with this new model we get the following results 0 04mM Protein Sample Normalized to either 0 02 0 04 or 0 08mM Fitting Model Includes Concentration Scalar Parameter 30 Normalized to 0 08mM 40 50 es lt 60 E 70 80 Normalized to 0 04mM wv 90 100 110 120 130 140 150 Normalized to 0 02mM 160 1 0 180 190 200 290 300 310 320 330 340 350 Temp K Parameter Normalized to Normalized to Normalized to 0 08mM 0 04mM 0 02mM Ao 0 0 0 Ay 0 0 0 Bo 305 4946027 305 4946039 305 494
20. l minimum for the refined parameters when the initial guesses are poor and second order polynomials for Cn T and Cp T are used This has been observed only rarely To avoid this potential problem always refine using linear baselines first then refine with second order polynomial functions if desired To confirm that the solution is not a local minimum when using second order polynomial functions perform several rounds of refinement switching back and forth between linear and second order functions for Cy T and Cp T VI Situations Where TD Does Not Exist Various definitions in the model depend upon a temperature value for TD Data can be analyzed under conditions where the protein is significantly destabilized However the data cannot be accurately analyzed if there is no temperature at which the native and denatured states are equally populated i e no TD 21 DSCFIT v 15 3 User s Manual 03 25 2002 http wine1 sb fsu edu DSCfTit Section 5 Derivations Native state and denatured state heat capacity functions Cy T 3 4 7 TD 2 3 7 TD C C T 3 A P FDY 2 B T TD DCTD C ACp T is ACT VSS AHA Vt 7D 228 8 8 tr SID s Dor Thus AH T can be written as ABTS A SA Sey A a Sy ec er eo AS T is l Y a AS T mcr ro 3 70 DE TEDS P parca ee 14 h 7D In 7 D F 2 5B B bn 7 b 7D DCT D oe POTD and AG AG AH T AS This treaT Dent now allows u
21. les calories 4 184 directly to the data when imported If a mistake is made with the units when importing the data close the project window and read in the data file again ll Initial Guess of Thermodynamic Parameters After reading in a data file DSCFIT will automatically determine an initial guess for the various thermodynamic parameters and report these values in the Parameters Dialog Box Set Parameters Remain Fixed lt l First order term for Native State Bol 296 150854640 ero Order term for Native State Co 128881 92707 Second order term for Denatured State AT jo First order term for Denatured State E1 31 3341814309 Delta Cp at Melting Temperature DETO 7329 00695677 Enthalpy at Melting Temperature POTO 266435 278331 Melting Temperature TO 212 602861960 Print Results Kea if m Eu E E lt DSCFIT v 15 3 User s Manual 03 25 2002 http wine1 sb fsu edu DSCfTit At this point you can simply click OK to accept the program s initial guesses for the parameters Alternatively you can modify any of the parameters Checking the left hand side box next to a parameter will constrain the parameter during refinement to the value indicated in the right hand side box Leaving the left hand side box unchecked will allow the value of the parameter to vary during refinement to improve the agreement of the model with the data Three parameters Ag the second order term for the
22. ma Ba m Ta t oA i 120 wan Ba L E u E B un L Ba 300 j0 200 300 210 320 330 340 l l l l I l l l l l Ready NUM A Non random distribution of the scatter can be indicative of non 2 state behavior A peak or trough centered at the TD is one of the features of a transition that is too broad or too narrow to be fit by a 2 state model 15 DSCFIT v 15 3 User s Manual 03 25 2002 http wine1 sb fsu edu DSCfTit re DSCFit DSCFitt File Edit View Functions Window Help Load Data Graphs Refine Evaluate Parameters Export Residual Scatter E 1620 ral 2440 290 300 310 320 330 340 NM 4 In this example the transition of the raw data is too broad for a 2 state model Thus the fit tends to overshoot the peak of the raw data and undershoot on either side of the peak In addition to the systematic nature of the residual scatter the magnitude of the error is greater compare to prior scatter plot Non 2 state behavior due to an intermediate state can also result in aberrant values for ACp T at the TD In particular apparent negative values for ACp at the TD i e the value of the DCTD parameter may result 16 DSCFIT v 15 3 User s Manual 03 25 2002 http wine1 sb fsu edu DSCtTit Set Parameters Remain Fised a second order term for Native State o fo First order term for Native State Bo 503 763807744 E ero Order term
23. native state heat capacity function A the second order term for the denatured state heat capacity and k the concentration constant are constrained as indicated by the appropriately checked box to values of 0 0 and 1 respectively The text descriptions of these parameters are also boxed to indicate that refinement of these parameters may be justified only under certain circumstances Ill Graphs Dialog Box After assignment of the initial values for the various thermodynamic parameters DSCFIT will display the Graphs Dialog Box Dialog Please Select Data Curves one ja Please Choose a Graph f Raw Data and Fit Curves IY Raw Data smar 346 Residual Scatter Plat If Heat Capacity CIT yrnirt e5000 C Delta Cp If Native State Cri T ymar 45000 C DeltaH Il Denatured State Co T F Delta G nf Auto Scale C Deltas The Graphs Dialog Box allows you to select the appropriate thermodynamic function s for viewing as a graph The graph chosen will be actively updated during the refinement cycle The available functions include Raw Data and Fit Curves this graph will include the raw data an can also include the functions for the native state heat capacity function denatured state heat capacity function as well as the fitted function based upon the current values of the various thermodynamic variables This is the default graphic representation Residual Scatter Plot this graph shows the disagreement between
24. ne1 sb fsu edu DSCtTit Although DSCFIT allows the use of second order polynomials during the refinement it is up to the user to decide if the use of a second order polynomial is valid The key issue is the accuracy of the baselines at temperatures far away from the TD The use of second order polynomial baselines can result in unreasonable behavior at extremes of temperature Thus while second order polynomial functions for Cp T or Cy T may improve accuracy at the TD the accuracy at other temperatures may be reduced IV Determination of Thermodynamic Constants at Temperatures other than TD It is often important to evaluate the values for the various thermodynamic constants at temperatures other than the TD For example determination of the effects of a destabilizing or stabilizing mutation requires that the AG value for the mutant be determined at the TD of the wild type protein to obtain a AAG value The value of the various thermodynamic parameters at any reference temperature can be obtained by using the Functions Evaluate Point menus or the Evaluate menu button This will bring up the Evaluate dialog box Evaluate Temperature 300 Raw Data j 3801910 5 O Heat Capacity 1 3766210 5 Delta Cp isezes0 4 Delta H 28257105 Delta G P3409910 3 OO ubrit Input the desired reference temperature in K in the Temperature box and click the Submit button The values for the various thermodynamic parameters at th
25. oceed to section 1 General Steps in Data Analysis General descriptions of thermodynamic parameters are found in Section 2 and derivations of the thermodynamics parameters are found in Section 5 Section 3 Available Models and Section 4 Advanced Topics provide information on the various features of the program and their application in data analysis DSCFIT v 15 3 User s Manual 03 25 2002 http wine1 sb fsu edu DSCfTit section 1 General Steps in Data Analysis Loading of Data The DSC data to be analyzed in DSCFIT should be a finalized molar heat capacity data file In other words it should be a protein buffer run that has had the buffer buffer run subtracted and then been normalized for the molar concentration of the protein although as discussed in the Advanced Concepts section DSCFIT can determine the apparent molar concentration with certain assumptions The data file should be a text file consisting of two columns the first column will contain the temperature data and the second column will contain the corresponding values for the molar heat capacity These columns of data may be separated by either a space or spaces or a tab character DSCFIT will not currently parse data delineated by other characters e g commas The temperature can be either in C or K and the molar heat capacity can be in either calorie or Joule units DSCFIT works internally with K and Joules however Load data by using either the File
26. owed to float if other data Is available to confirm either the concentration or the value of AHyr AHea See Advanced Topics section Section 4 Advanced Topics Data Clipping The range of data to be analyzed can be defined using both a high and low temperature cutoff value The first step is to identify the temperature range s for exclusion Place the mouse pointer at the appropriate position in any of the displayed graphs and click the right mouse button to obtain the x y coordinates of that point DSCFIT v 15 3 User s Manual 03 25 2002 http wine1 sb fsu edu DSCTi 2 DSCFit DSCFit1 File Edit View Functions Window Help x 7 Load Data Graphs Refine Evaluate Parameters Export Heat Capacity Graphs wi a 102600 e ar ae f p 113400 ai Coordinate Values x Sets el n T 338 61 value 1 1655e 005 124200 y i 135000 fv ae 145600 ae z 290 300 310 320 330 340 Pai i l i Ready NUM A Once the specific high and low temperature cutoff values have been identified you can limit the data range used during refinement by selecting the Functions Constrain Data options from the menu bar e DSCFit DSCFitl File Edit Mew Emmie vindow Help Lo Graphs Refine Evaluate Parameters Refine Be DSCFitl Evaluate Point Parameters Constrain Data 30323 This will bring up the Constrain Data Dialog box 12 DSCFIT v 15 3
27. propriate to refine with the value for DCTD constrained to the known value The nature of the ACp T function is defined by the Cy T and Cp T functions i e ACp T Cp T Cn T Thus if both Cn T and Cp T functions are linear then ACp T will be a linear function etc DSCFIT v 15 3 User s Manual 03 25 2002 http wine1 sb fsu edu DSCfTit IV DDTD and TD If desired the value of AH T at the TD can be constrained Also the value of TD can be constrained V k As discussed in the Description of Thermodynamic Parameters the k parameter is a scalar term applied to the raw data molar heat capacity Investigators can use the refined value of k to gauge potential errors in concentration and inequality of AHy4 and AHea If it is allowed to float during refinement and does not refine to a value of 1 0 there are two possible interpretations 1 AHw AHea 1 0 but the concentration applied during normalization of the molar heat capacity is incorrect In this case the actual concentration of the protein in the DSC analysis is the value used during the normalization k Furthermore the refined thermodynamic parameters are for the corrected concentration 2 The concentration is correct but AHyy is not equal to AHga In this case the value of K is equal to AHyy AH a Furthermore the refined value of DDTD is equal to AHyy AHca can be derived by dividing DDTD van t Hoff enthalpy by k Thus k should only be all
28. rations since last change f 0 Coefficients E mit Standard Deviation values around 100 J mol K represent an excellent fit with 2 state behavior Typical values range around 100 500 J mol K Values above 500 J mol K may indicate non 2 state behavior Another property to evaluate is the residual scatter of the data and to decide whether is it random or systematic The residual scatter of the data is a graphical representation of the difference between the fitted model and the raw data fit raw data Select the Residual Scatter Plot radio button from the Graphs dialog box selected from the Graphs menu button Dialog Please Select Data Curves E Re Dieta i Pea Capacity Ei m Wetive state Ennii M Denatured State Edit min Pa Please Choose a Graph C Raw Data and Fit Curves ymas 346 Delta Cp C DeltaH C Delta G Deltas T ymin 153291 pmax f 985346 W Auto Scale A graph of the residual scatter will be displayed This graph will include a vertical green line to indicate the location of the melting temperature TD as well as a blue horizontal line indicating the zero location of the scatter 14 DSCFIT v 15 3 User s Manual 03 25 2002 http wine1 sb fsu edu DSCfTit J DSCFit DSCFit1 File Edit View Functions Window Help _ x Load Data Graphs Petine Evaluate Parameters Export Residual Scatter Tas a m pr a a n a ol Pa
29. s to describe the model function using the following parameters Ao Bo Co A1 B4 DOTD DDTD TD Where Ao Bo Co are terms for the second order polynomial describing the native state heat capacity function A B are the first two terms for the second order polynomial describing the denatured state heat capacity function DCTD is the value of AC at the melting temperature DDTD is the enthalpy of the system at the melting temperature and TD is the melting temperature This allows us to independently define the native and denatured state baselines as being either second order polynomial functions linear functions or some constant value It also allows us to set ACp at the melting temperature i e DCTD and AH at the melting temperature i e DDTD to some fixed value during parameter refinement Thus we now have a very flexible model with complete control over all critical parameters during refinement The fractional component of native state as a function of AG is equal to 22 DSCFIT v 15 3 User s Manual 03 25 2002 http wine1 sb fsu edu DSCfit By T i ol al By knowing the native state heat capacity function Cy T the difference heat capacity function between the denatured and native states ACp the fractional component of native state as a function of temperature and the enthalpy function the heat capacity of the system as a function of temperature can be determined r fpj 1 F OD Op T AC
30. ts for the Cy T and Cp T functions as well as the ACp T and AH T functions l Ca Cn T can be modeled as either a second order polynomial linear function or constant independent of the treatment of the Cp T function Ao is non zero Cn T is a second order polynomial Ao 0 Cy T is a linear function slope Bo y intercept Co Ao and Bo 0 Cn T is a constant equal to Co Values for Ap Bo and Co can be allowed to float during refinement or may be fixed at some operator defined value the model will find the best function with these applied constraints Il Cp T Like Cy T Cp T can be modeled as either a second order polynomial linear function or constant independent of the treatment of the Cy T function A is non zero Cp T is a second order polynomial A 0 Cp T is a linear function slope B4 y intercept C4 A and B 0 Cp T is a constant equal to DC TD Co Values for A and B can be allowed to float during refinement or may be fixed at some operator defined value the model will find the best function with these applied constraints Il DOTD DCTD defines the value of ACp T at the TD This value can be fixed or allowed to float during refinement Of all the thermodynamic parameters this is the most difficult value to determine accurately and is sensitive to errors in concentration determination In some cases i e where accurate values for ACp T at the TD are known it may be ap
Download Pdf Manuals
Related Search