Comparison of logP and logD correction models trained with public and proprietary data sets

Aliagas, Ignacio; Gobbi, A.; Lee, Man-Ling; Sellers, Benjamin D.

doi:10.1007/s10822-022-00450-9

Cited by 16 publications

(20 citation statements)

References 51 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Experimental uncertainty range compiled from literature sources is between <0.2 and 0.4 log units. 23 , 29 , 30 Note the logarithmic scale of both axes. All error bars are smaller than the plot markers and have been omitted for visual clarity.…”

Section: Resultsmentioning

confidence: 99%

“…26 Experimental uncertainty estimates for log P measurements range from <0.2 to 0.4 log units and depend on the specific experimental setup. 23,29,30 However, both direct and indirect methods struggle to measure log P for poorly soluble, very lipophilic (high log P), or very hydrophilic (low log P) compounds. 26 This is wellillustrated by the relatively large fraction of compounds in the in-house data set with a measured log P annotated as "> 4.7" (Figure S1), highlighting the limits of the used experimental method (see the Methods section).…”

Section: ■ Introductionmentioning

confidence: 99%

“… 26 Experimental uncertainty estimates for log P measurements range from <0.2 to 0.4 log units and depend on the specific experimental setup. 23 , 29 , 30 …”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Machine Learning for Fast, Quantum Mechanics-Based Approximation of Drug Lipophilicity

et al. 2023

View full text Add to dashboard Cite

Lipophilicity, as measured by the partition coefficient between octanol and water (log P), is a key parameter in early drug discovery research. However, measuring log P experimentally is difficult for specific compounds and log P ranges. The resulting lack of reliable experimental data impedes development of accurate in silico models for such compounds. In certain discovery projects at Novartis focused on such compounds, a quantum mechanics (QM)-based tool for log P estimation has emerged as a valuable supplement to experimental measurements and as a preferred alternative to existing empirical models. However, this QM-based approach incurs a substantial computational cost, limiting its applicability to small series and prohibiting quick, interactive ideation. This work explores a set of machine learning models (Random Forest, Lasso, XGBoost, Chemprop, and Chemprop3D) to learn calculated log P values on both a public data set and an in-house data set to obtain a computationally affordable, QM-based estimation of drug lipophilicity. The message-passing neural network model Chemprop emerged as the best performing model with mean absolute errors of 0.44 and 0.34 log units for scaffold split test sets of the public and in-house data sets, respectively. Analysis of learning curves suggests that a further decrease in the test set error can be achieved by increasing the training set size. While models directly trained on experimental data perform better at approximating experimentally determined log P values than models trained on calculated values, we discuss the potential advantages of using calculated log P values going beyond the limits of experimental quantitation. We analyze the impact of the data set splitting strategy and gain insights into model failure modes. Potential use cases for the presented models include pre-screening of large compound collections and prioritization of compounds for full QM calculations.

show abstract

Section: Resultsmentioning

confidence: 99%

Section: ■ Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Machine Learning for Fast, Quantum Mechanics-Based Approximation of Drug Lipophilicity

et al. 2023

View full text Add to dashboard Cite

show abstract

“… a Inhibition of the NLRP3 inflammasome pathway was assessed by quantifying IL-1β release from lipopolysaccharide (LPS) + nigericin stimulated hPBMCs ( n = 1 for all compounds except compound 1 which is reported as the geometric mean ± SD). b LLE was calculated using the following equation: LLE = pIC 50 IL-1β – cLogD. , …”

Section: Resultsmentioning

confidence: 99%

“…b LLE was calculated using the following equation: LLE = pIC 50 IL-1β − cLogD. 25,26 aryl-S connection were detrimental to potency (15); however, we discovered that the potency could be restored by cyclizing the two pyrazole substituents to form the fused bicyclic pyrazole 16. We were encouraged to see that further improvements to potency and LLE could be made by introducing an oxygen atom into the saturated fused ring system to generate the 3-substituted-6,7-dihydro-5H-pyrazolo-[5,1-b][1,3]oxazine (17).…”

Section: ■ Introductionmentioning

confidence: 99%

Overcoming Preclinical Safety Obstacles to Discover (S)-N-((1,2,3,5,6,7-Hexahydro-s-indacen-4-yl)carbamoyl)-6-(methylamino)-6,7-dihydro-5H-pyrazolo[5,1-b][1,3]oxazine-3-sulfonamide (GDC-2394): A Potent and Selective NLRP3 Inhibitor

McBride¹,

Trzoss²,

Povero³

et al. 2022

J. Med. Chem.

View full text Add to dashboard Cite

Inappropriate activation of the NLRP3 inflammasome has been implicated in multiple inflammatory and autoimmune diseases. Herein, we aimed to develop novel NLRP3 inhibitors that could minimize the risk of drug-induced liver injury. Lipophilic ligand efficiency was used as a guiding metric to identify a series of 6,7-dihydro-5H-pyrazolo[5,1-b][1,3]oxazinesulfonylureas. A leading compound from this series was advanced into safety studies in cynomolgus monkeys, and renal toxicity, due to compound precipitation, was observed. To overcome this obstacle, we focused on improving the solubility of our compounds, specifically by introducing basic amine substituents into the scaffold. This led to the identification of GDC-2394, a potent and selective NLRP3 inhibitor, with an in vitro and in vivo safety profile suitable for advancement into human clinical trials.

show abstract

Neural Network Models for Predicting Solubility and Metabolism Class of Drugs in the Biopharmaceutics Drug Disposition Classification System (BDDCS)

Ashrafi,

Teimouri,

Aghazadeh

et al. 2023

Eur J Drug Metab Pharmacokinet

View full text Add to dashboard Cite

Comparison of logP and logD correction models trained with public and proprietary data sets

Cited by 16 publications

References 51 publications

Machine Learning for Fast, Quantum Mechanics-Based Approximation of Drug Lipophilicity

Machine Learning for Fast, Quantum Mechanics-Based Approximation of Drug Lipophilicity

Overcoming Preclinical Safety Obstacles to Discover (S)-N-((1,2,3,5,6,7-Hexahydro-s-indacen-4-yl)carbamoyl)-6-(methylamino)-6,7-dihydro-5H-pyrazolo[5,1-b][1,3]oxazine-3-sulfonamide (GDC-2394): A Potent and Selective NLRP3 Inhibitor

Neural Network Models for Predicting Solubility and Metabolism Class of Drugs in the Biopharmaceutics Drug Disposition Classification System (BDDCS)

Contact Info

Product

Resources

About