Dataset profile

The two available at OCHEM models predict Melting Point (MP) of organic chemical compounds. The MP is one of the important physico-chemical properties, which is frequently used in drug discovery to determine estimate aqueous solubility of chemical compounds. The complexity with prediction of this point are connected to purity of compounds, existence of polymorphic forms, degradation of compounds before melting, etc. All these factors influence the quality of models for this point. The data for MP were collected in OCHEM database database  as well as were provided by Dr. Luc Patiny from ChemExper database. ChemExper database (OCHEM dataset) and Enamine Ltd (Enamine dataset) The majority of data were organic chemistry compounds. The models were validated using 277 compounds compiled by [Bergstrom et al  Molecular descriptors influencing melting point and their role in classification of solid drugs. J. Chem. Inf. Comput. Sci. 2003; 43 (4) 1177-85] as well as data from Open Notebook.

Data preprocessing

All chemical structures were processed using  OCHEM cleaning and standardization protocols. A specific care was used to eliminate salts and mixtures, and inorganic compound,  which could dramatically change MP of molecules. The detection and elimination of outliers was done usingbased on p-value (article in preparation).

Descriptors

The first model (2D) was built using a combination of ALOGPS 2.1 model predictions and EState descriptors (electrotopological EState indices). The EState indices were calculated using a program developed by Dr. Tanchuk, which was also used to develop ALOGPS 2.1 model.

The second model (3D) was built using DRAGON descriptors, which was provided by Prof. Todeschini and Talete Srl. For this model we generated 3D conformations of molecules using CORINA software, which is distributed by Molecular Networks GmbH.

Validation

The model was Models were built 11 individual descriptor packages available in OCHEM. A simple average of all 10 models was done to develop consensus model. This model, however, requires rather long calculations, especially if calculations of descriptors have not been previously cached. There is also 2D model,  "Melting Point best (Estate)", which  was built using Estate descriptors. All other sub-models are not shown, just to avoid confusion with having too many of them in the web browser (they can be  accessed using public IDs indicated in the profile of the consensus model as https://ochem.eu/model/MODEID).

Validation

The model were built using 5-fold cross validation . The dataset of 277 compounds compiled by [Bergstrom et al  ]as well as prediction of subsets (e..g, model developed using OCHEM subset was used to predict Enamine, Bergstrom and Bradley sets, etc.)

Statistical parameters

Prediction accuracy

The basic prediction accuracy parameters according to the 5-fold cross-validation procedure (N=25547) are:

0.300.52

Property
	# records	RMSE	MAE	R2	r2 (Coefficient of determination)
logP	16912	0.42	Consensus model	37.1	27.6	0.9578	0.9578
logS	8102	0.70	Melting Point best (Estate)	39.6	29.1	0.9075	0.9075

Applicability domain

The prediction accuracy is estimated using ASNN-STD. This distance to model  was shown to provide the best assessment of the accuracy of prediction as described in [Tetko et al, Critical assessment of QSAR models of environmental toxicity against Tetrahymena pyriformis: focusing on applicability domain and overfitting by variable selection, J Chem Inf Model. 2008 Sep;48(9):1733-46accuracy for drug-like subset (molecules with melting point in [50,250]°C interval is less than 33°C for the consensus model.

Reference

The full details of the study are published in How accurately can we predict the melting points of drug-like compounds? [Tetko IV, Sushko Y, Novotarskyi S, Patiny L, Kondratov I, Petrenko AE, Charochkina L, Asiri AM. J Chem Inf Model. 2014 Dec 22;54(12):3320-9. doi: 10.1021/ci800151m].ci5005288.]

Availability

All data can be publicly downloaded at http://ochem.eu/article/55638.