Overall accuracy

As examples of applications, we present the overall accuracy of predicted ionization constants for about 50 groups in 4 proteins, changes in the average charge of bovine pancreatic trypsin inhibitor at pH 7 along a molecular dynamics trajectory, and finally, we discuss some preliminary results obtained for protein kinases and protein phosphatases. 

The salient comparisons are between the bars marked P3-Dk, our initial parallel PME implementation, and DP-4, the macroscopic multipole method with four levels of macroscopic boxes. Though it is difficult to create a completely fair comparison in terms of the relative accuracy of the potentials and forces as computed by the two methods, the parameters for these simulations were tuned to give comparable overall accuracy. PME is clearly... 

As mentioned earlier, overall accuracy of finite element computations is directly detennined by the accuracy of the method employed to obtain the numerical solution of the global system of algebraic equations. In practical simulations, therefore, computational errors which are liable to affect the solution of global stiffness equations should be carefully analysed. 

Critica.1 Properties. Several methods have been developed to estimate critical pressure, temperature, and volume, U). Many other properties can be estimated from these properties. Error propagation can be large for physical property estimations based on critical properties from group contribution methods. Thus sensitivity analyses are recommended. The Ambrose method (185) was found to be more accurate (186) than the Lyderson (187) method, although it is computationally more complex. The Joback and Reid method (188) is only slightly less accurate overall than the Ambrose method, and is more accurate for some specific substances. Other methods of lesser overall accuracy are also available (189,190) (T, (191,192) (T, P ),... 

Computational accuracy can be dramatically improved by dynamically adding elements where they minimize the error. For example, more elements ean be added in the neighborhood of a strong gradient in the velocity to help resolve shocks and vortex sheets. Elements may be removed from regions of smooth flow to minimize the computational cost without degrading the overall accuracy. The concept is shown in Fig. 9.6 where a finer mesh overlays the original mesh. This mesh refinement can be carried out to as many levels as necessary [15], [16], [17]. 

Models are arranged in the plot in order of decreasing MAD, and thus decreasi overall accuracy. 

Wigner s study of the correlation energy in the alkali metals was a pioneering work of great importance, and the degree of difficulty of this field is perhaps best demonstrated by the fact that, during the last two decades, little improvement has been achieved in his correlation formula III. 12. The formula is nevertheless rather crude, and Wigner himself estimates the overall accuracy to be within 20 per cent. 

Minute amounts of coelenterazine can also be measured utilizing apoaequorin or apoobelin (Campbell and Herring, 1990 Thompson et ah, 1995). In this method, a sample containing coelenterazine is treated with an excess amount of apophotoprotein (apoaequorin or apoobelin) to convert it to a Ca2+-sensitive photoprotein (aequorin or obelin). The photoprotein formed is assayed by luminescing it with Ca2+ to determine the amount of coelenterazine originally existed. With this method, the luminescence reaction is fast and usually complete in a few seconds, in contrast to the slower luminescence reactions with luciferases that sometimes require a few minutes to complete. However, the formation of photoprotein from apoaequorin is slow and not necessarily quantitative, and the overall accuracy of the photoprotein method does not compare favorably with that of the luciferase method that directly measures coelenterazine. The author recommends using a luciferase if the enzyme is available. 

In the isoenthalpic and isoentropic reaction series, the isoenthalpic and isoentropic activation parameters, respectively, may be defined as a special case of isokinetic values. Values of this kind are certainly more reliable than the usual ones in all cases when the isokinetic behavior has been proved, or, better, when it cannot be rejected. In dubious cases, when it cannot be disproved with any reliability but yet would cause a reduction of the overall accuracy, we recommend the calculation of both isokinetic and unconstrained parameters. In... 

To provide a measure of the state of the art for a test. The standard deviation is a measure of the ability of different laboratories to obtain the same result. Where the true concentration is known the results show overall accuracy. 

Having these severe approximations in mind, SCC-DFTB performs surprisingly well for many systems of interest, as discussed above. However, it has a lower overall accuracy than DFT or post HF methods. Therefore, applying it to new classes of systems should be only done after careful examination of its performance. This can be done e.g. by conducting reference calculations on smaller model systems with DFT or ab initio methods. A second source of errors is related to some intrinsic problems with the GGA functionals also used in popular DFT methods (SCC-DFTB uses the PBE functional), which are inherited in SCC-DFTB. This concerns the well known GGA problems in describing van der Waals interactions [32], extended conjugate n systems [45,46] or charge transfer excited states [47, 48],... 

The aptness of the idealized sd/J Lewis-like model is also confirmed by the quantitative NBO descriptors, as summarized inTable4.5. This table displays the overall accuracy of the Lewis-like description (in terms of %pl, the percentage accuracy of the natural Lewis-like wavefunction for both valence-shell and total electron density) as well as the metal hybridization (hM), bond polarity toward M (100cm2), and... 

Wajima and coauthors offer an alternative approach to utilize animal VD data to predict human VD [13]. Several compound descriptors that included both chemical structural elements as well as animal VDSS values were subject to multiple linear regression and partial least squares statistical analyses, with human VDSS as the independent parameter to be predicted using a dataset of 64 drugs. Methods derived in this manner were compared to simple allometry for overall accuracy. Their analyses yielded the following regressions ... 

The overall accuracy of the predictions, assessed as the mean-fold error of prediction of the test set was 2.03, making this approach one that would possess suitable accuracy for use in drug design and human pharmacokinetic predictions. Similar methods developed separately for acids and bases showed an improvement in accuracy. This investigation also included a prediction of unbound VD, which should represent a simpler parameter to predict since it would be based only on tissue binding and not plasma protein binding. However, it is interesting to note that this approach was less accurate for this parameter, which would be unexpected. 

Overall accuracy 86% with zero hydrogen atoms and high number of the O-C hydrides. P-gp substrate showed MolES > 110 ... 

P-gp substrate 116 substrates and 85 nonsubstrates 19 substrates and 6 nonsubstrates 84.2% substrates and 66.7% nonsubstrates correctly predicted. Overall accuracy 86%... 

P-gp substrate 22 substrates and 31 nonsubstrates 115 substrates and 157 nonsubstrates 61% substrates and 81% nonsubstrates correctly predicted. Overall accuracy 72.4% Transport Caco-2 cell line Size, shape (e.g. molecular surface and glo-bularity), hydrophilic and H-bonding related descriptors correlated positively with P-gp activity, log P0/w not significant [54]... 

---