Skip to main content

Outputs

Location​

The results of the fits are written in the directory Output/ by default. However, you can change this location using the command-line option --output (or -o) followed by the desired path name.

chemex fit -o <path> [other options]

Multi-step fit​

If the fitting method has multiple fitting steps, each step will create its own output subdirectory with the name of the fitting step.

Output/
├── STEP1/
└── STEP2/

Multi-group fit​

During any fitting step, if the dataset can be divided in multiple groups depending on distinct sets of fitting parameters, ChemEx will fit each group of data separately. The results of these individual fits are then stored in separate subfolders placed in the the Groups/ folder. Subfolders are named with a number and an ID, which depends on the parameters that have been optimized for the specific group. The results are also put together in the All/ folder for convenience.

For example, if all global parameters (pB, kex, etc.) are fixed or residue-specific fitting model is used (e.g. the model 2st_rs), then all parameters are allowed to vary are residue-specific. Residue-specific fits will then be run and two separate subdirectories All/ and Groups/ will be created, which contain fitting results for all residues and each individual residue, respectively.

Output/
└── STEP2/
    ├── All/
    └── Groups/
        ├── 10_11N/
        ├── 11_12N/
        ├── 12_13N/

Content​

The fitting output typically contains the following files and directories:

Parameters/​

Contains fitting results as three separate files fitted.toml, fixed.toml and constrained.toml, which contain output parameters that are fitted, fixed and constrained during the fitting process, respectively.

Example files​

[GLOBAL]
KEX_AB =  3.81511e+02 # ±8.90870e+00
PB     =  7.02971e-02 # ±1.14784e-03

[DW_AB]
15N =  2.00075e+00 # ±2.30817e-02
31N =  1.98968e+00 # ±1.90842e-02
33N =  1.82003e+00 # ±2.44821e-02
34N =  3.63170e+00 # ±3.62801e-02
37N =  1.69183e+00 # ±2.41070e-02

["R2_A, B0->500.0MHZ"]
15N =  3.98674e+00 # ±2.55793e-01
31N =  5.85923e+00 # ±1.94734e-01
33N =  4.02099e+00 # ±2.78003e-01
34N =  4.16615e+00 # ±2.05190e-01
37N =  3.67705e+00 # ±3.04621e-01

["R2_A, B0->800.0MHZ"]
15N =  6.33712e+00 # ±3.86894e-01
31N =  7.99927e+00 # ±2.81972e-01
33N =  6.23967e+00 # ±5.82366e-01
34N =  5.99535e+00 # ±3.17832e-01
37N =  5.37295e+00 # ±5.69929e-01
note

If uncertainties on fitted parameters have been calculated – this depends on the fitting algorithm used –, they are reported as comments at the end of the line and are preceded by the sign "±". Uncertainties on fitted parameters are, typically, estimated through the covariance matrix obtained from the Levenberg-Marquardt optimization.

Plot/​

Contains fitting results as plots (in .pdf format) and also the raw datasets (both the original input and fitted data points) for creating the plots. Example fitting results for CPMG and CEST experiments are shown below:

CEST profileCPMG profile
Examples of CEST and CPMG fitting results
note

In plots of (D-/cos-)CEST fitting results, the positions for ground and excited states are indicated by solid and dashed vertical lines respectively. Besides, data points that are filtered out from the fit are shown in lighter color. In plots of D-CEST/COS-CEST fitting results, the "folded" positions for ground and excited states are indicated by "*" at the vertical lines.

Data/​

Contains all the data values used for the fitting along with the back-calculated values. These files can be used to calculate the χ2χ^2 value.

[15N]
#         NCYC   INTENSITY (EXP)       ERROR (EXP)  INTENSITY (CALC)
             0    3.47059800e+04    1.77491406e+02    3.47055362e+04
            30    3.05930380e+04    1.77491406e+02    3.06111963e+04
             1    1.81234230e+04    1.77491406e+02    1.80856300e+04
            28    3.07144730e+04    1.77491406e+02    3.05838767e+04
             2    2.02155120e+04    1.77491406e+02    2.02110184e+04
            26    3.05222020e+04    1.77491406e+02    3.05501255e+04
             3    2.23056070e+04    1.77491406e+02    2.23601656e+04
            24    3.05381830e+04    1.77491406e+02    3.05077686e+04
             4    2.43783050e+04    1.77491406e+02    2.44111932e+04
            22    3.06981570e+04    1.77491406e+02    3.04536377e+04
             5    2.58673980e+04    1.77491406e+02    2.59884541e+04
            20    3.06069180e+04    1.77491406e+02    3.03829761e+04
             6    2.70660870e+04    1.77491406e+02    2.71148924e+04
            18    3.00982700e+04    1.77491406e+02    3.02883916e+04
             7    2.81512990e+04    1.77491406e+02    2.79178594e+04
            16    3.02515700e+04    1.77491406e+02    3.01579211e+04
             8    2.84045570e+04    1.77491406e+02    2.84985729e+04
            14    2.97528530e+04    1.77491406e+02    2.99712519e+04
             9    2.84536650e+04    1.77491406e+02    2.89264158e+04
            13    2.98219180e+04    1.77491406e+02    2.98461022e+04
            10    2.96936410e+04    1.77491406e+02    2.92495591e+04
            12    2.95269900e+04    1.77491406e+02    2.96918706e+04
            11    2.92540210e+04    1.77491406e+02    2.94972506e+04
             2    2.04476470e+04    1.77491406e+02    2.02110184e+04
            28    3.05466550e+04    1.77491406e+02    3.05838767e+04
             8    2.86183900e+04    1.77491406e+02    2.84985729e+04

[31N]
#         NCYC   INTENSITY (EXP)       ERROR (EXP)  INTENSITY (CALC)
             0    4.71577550e+04    1.77491406e+02    4.71537007e+04
            30    4.00023250e+04    1.77491406e+02    3.94823953e+04
             1    2.30167260e+04    1.77491406e+02    2.34136180e+04
            28    4.00615990e+04    1.77491406e+02    3.94547546e+04
             2    2.61934530e+04    1.77491406e+02    2.61653780e+04
            26    3.96898190e+04    1.77491406e+02    3.94190638e+04
             3    2.86882570e+04    1.77491406e+02    2.88729551e+04
            24    3.98238690e+04    1.77491406e+02    3.93722573e+04
             4    3.21031440e+04    1.77491406e+02    3.13942278e+04
            22    3.96733310e+04    1.77491406e+02    3.93097577e+04
             5    3.35871430e+04    1.77491406e+02    3.34628362e+04
            20    3.86957600e+04    1.77491406e+02    3.92245398e+04
             6    3.53600830e+04    1.77491406e+02    3.49641428e+04
            18    3.88115960e+04    1.77491406e+02    3.91054958e+04
             7    3.58938150e+04    1.77491406e+02    3.59453347e+04
            16    3.85843960e+04    1.77491406e+02    3.89345050e+04
             8    3.61450060e+04    1.77491406e+02    3.66900206e+04
            14    3.85711880e+04    1.77491406e+02    3.86811094e+04
             9    3.72815770e+04    1.77491406e+02    3.72429094e+04
            13    3.82358560e+04    1.77491406e+02    3.85092366e+04
            10    3.76426640e+04    1.77491406e+02    3.76803042e+04
            12    3.78470460e+04    1.77491406e+02    3.82929922e+04
            11    3.76000060e+04    1.77491406e+02    3.80220076e+04
             2    2.62301400e+04    1.77491406e+02    2.61653780e+04
            28    3.97731200e+04    1.77491406e+02    3.94547546e+04
             8    3.63510140e+04    1.77491406e+02    3.66900206e+04

[...]

statistics.toml​

Contains all goodness-of-fit statistics, such as χ2χ^2.

"statistics.toml"
"number of data points"                = 230
"number of variables"                  = 17
"chi-square"                           =  4.34824e+02
"reduced-chi-square"                   =  2.04143e+00
"chi-squared test"                     =  0.00000e+00
"Kolmogorov-Smirnov test"              =  6.72236e-02
"Akaike Information Criterion (AIC)"   =  1.80479e+02
"Bayesian Information Criterion (BIC)" =  2.38926e+02