Upload
stephen-stewart
View
217
Download
1
Tags:
Embed Size (px)
Citation preview
STUDY OF THE SECOND VIRIAL COEFFICIENTS:NEW CHALLENGE FOR QSPR
Elena Mokshyna, Victor E. Kuz’min, Vadim I. Nedostup
WHY CHALLENGE?
The compressibility factor is expressed as a series expansion in either density (reciprocal molar volume) or pressure:
Main purposes:
• Development of approach to QSPR of T-dependent properties• Calibration of the descriptors
• Prediction for new complex organic compounds
EXPERIMENTAL DATA
Number of compounds: 262 Number of points: 4787
Range of virial coefficients: -5891 – 391 cm3/mol Range of temperatures: 110 – 773 K
DESCRIPTORS & MODELLING TECHNIQUES
SiRMS descriptors:
Temperature as a single descriptor:
B = f(T)
Two-layer QSPR model:
a = f(descriptors)
b = f(descriptors)
B = f(a, b)
STATISTICAL ANALYSIS
Various statistical methods:
MLR (Multi-Linear Regression)
PLS (Projection on Latent Structures)
RF (Random Forest)
SVM (Support Vector Machines) with radial basis function kernel
Rigorous 3x5-fold stratified external cross-validationTraining setTest set! Data on virial coefficient of compound
under all the temperatures are put in the test set
RESULTS for B = f(T)
R2ws = 0.53
R2ts = 0.19
R2ws = 0.71
R2ts = 0.45
R2ws = 0.87
R2ts = 0.68
R2ws = 0.94
R2ts = 0.71
RESULTS for B = f (a,b)a = f(descriptors), b = f(descriptors)
R2ws = 0.88
R2ts = 0.51
R2ws = 0.90
R2ts = 0.72
R2ws = 0.98
R2ts = 0.85
R2ws = 0.95
R2ts = 0.75
EXPERIMENTAL ERRORS VS. ERRORS OF MODELS
Hydro-carbons
Halocarbon compounds
Nitrogen compounds
Oxygen compounds
Silicon compounds
Sulphur compounds
0
40
80
120
160
200
Hydrocarbons; Exper-imental error; 21
Halocarbon com-pounds; Experimental
error; 24
Nitrogen compounds; Experimental error; 58
Oxygen compounds; Experimental error; 59Silicon compounds;
Experimental error; 50Sulphur compounds;
Experimental error; 48
Hydrocarbons; T-model error ; 59Halocarbon
compounds; T-model error ; 43
Nitrogen compounds; T-model error ; 190
Oxygen compounds; T-model error ; 131
Silicon compounds; T-model error ; 100
Sulphur compounds; T-model error ; 134
Hydrocarbons; Coef-ficient model error;
37
Halocarbon com-pounds; Coefficient
model error; 43
Nitrogen compounds; Coefficient model er-
ror; 172
Oxygen compounds; Coefficient model
error; 99Silicon compounds; Coefficient model
error; 75
Sulphur compounds; Coefficient model
error; 87Experimental errorT-model error Coefficient model error
Relative Variable Influence
Van-der-Waals Interactions
Temperature Partial charges Molecular Weight Donor/acceptor of Hydrogen Bond
0
10
20
30
40
Van-der-Waals In-teractions; Series1;
31
Temperature; Series1; 24
Partial charges; Series1; 18 Molecular Weight;
Series1; 16Donor/acceptor of Hydrogen Bond;
Series1; 11
Influential fragments
*
*
*
*
*
Some examples from the generated fragments library :
So….MISSION IS POSSIBLE,
BUT CHALLENGE IS NOT COMPLETED!
Thank you for the attention!