论文信息 - Non-linear Prediction of Quantitative Structure – Activity Relationships

Non-linear Prediction of Quantitative Structure – Activity Relationships

Predicting the log of the partition coefficient P is a long-standing benchmark problem in Quantitative Structure-Activity Relationships (QSAR). In this paper we show that a relatively simple molecular representation (using 14 variables) can be combined with leading edge machine learning algorithms to predict logP on new compounds more accurately than existing benchmark algorithms which use complex molecular representations.

I. Nabney | P. Tiňo | Yi Sun | B. S. Williams | J. Lösel

[1] A. Leo,et al. Substituent constants for correlation analysis. , 1977, Journal of medicinal chemistry.

[2] A. Leo,et al. Substituent constants for correlation analysis in chemistry and biology , 1979 .

[3] M. V. Rossum,et al. In Neural Computation , 2022 .

[4] H. White,et al. Universal approximation using feedforward networks with non-sigmoid hidden layer activation functions , 1989, International 1989 Joint Conference on Neural Networks.

[5] Heekuck Oh,et al. Neural Networks for Pattern Recognition , 1993, Adv. Comput..

[6] D. Signorini,et al. Neural networks , 1995, The Lancet.

[7] Geoffrey E. Hinton,et al. Bayesian Learning for Neural Networks , 1995 .

[8] Luís Torgo,et al. Functional Models for Regression Tree Leaves , 1997, ICML.

[9] J. Sangster. Octanol-Water Partition Coefficients: Fundamentals and Physical Chemistry , 1997 .

[10] Christopher K. I. Williams. Prediction with Gaussian Processes: From Linear Regression to Linear Prediction and Beyond , 1999, Learning in Graphical Models.

[11] L. Torgo,et al. Inductive learning of tree-based regression models , 1999 .