Low-Quality Structural and Interaction Data Improves Binding Affinity Prediction via Random Forest

Hongjian Li (1) , Kwong-Sak Leung (1) , Man-Hon Wong (1) , Pedro J. Ballester (2)

1 Department of Computer Science and Engineering
2 CRCM - Centre de Recherche en Cancérologie de Marseille

Docking scoring functions can be used to predict the strength of protein-ligand binding. It is widely believed that training a scoring function with low-quality data is detrimental for its predictive performance. Nevertheless, there is a surprising lack of systematic validation experiments in support of this hypothesis. In this study, we investigated to which extent training a scoring function with data containing low-quality structural and binding data is detrimental for predictive performance. We actually found that low-quality data is not only non-detrimental, but beneficial for the predictive performance of machine-learning scoring functions, though the improvement is less important than that coming from high-quality data. Furthermore, we observed that classical scoring functions are not able to effectively exploit data beyond an early threshold, regardless of its quality. This demonstrates that exploiting a larger data volume is more important for the performance of machine-learning scoring functions than restricting to a smaller set of higher data quality.

Mots clés: VALIDATION; SCORING FUNCTIONS; DATA SETS; IDENTIFICATION; DRUG DISCOVERY; ACCURACY; DOCKING

Domaines: Cancer

Hongjian Li, Kwong-Sak Leung, Man-Hon Wong, Pedro J. Ballester. Low-Quality Structural and Interaction Data Improves Binding Affinity Prediction via Random Forest. Molecules, 2015, ⟨10.3390/molecules200610947⟩. ⟨hal-01205333⟩