D3R Grand Challenge 4: Blind Prediction of Protein-Ligand Poses, Affinity Rankings, and Relative Binding Free Energies

parks, conor; Gaieb, Zied; Chiu, Michael; Yang, Huanwang; Shao, Chenghua; Walters, W. Patrick; Jansen, Johanna M.; McGaughey, Georgia B.; Lewis, Richard A.; Bembenek, Scott D.; Ameriks, Michael K.; Mirzadegan, Taraneh; Burley, S.K.; Amaro, Rommie E.; Gilson, Michael K.

doi:10.26434/chemrxiv.11363006.v1

Cited by 15 publications

(30 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…These results show that the performance of both models suffer on the GC4 dataset relative to the ChEMBL25 validation set. However, the performance is in line with the top performing ML models in prior D3R Grand Challenges (Gathiaka et al, 2016 ; Gaieb et al, 2018 , 2019 ; Parks et al, 2019a ). Finally, the prediction intervals from the conformal predictors are shown to be valid on the GC4 data set, demonstrating the validity of conformal prediction confidence intervals on a high-quality external test set.…”

Section: Introductionsupporting

confidence: 67%

“…We demonstrate that the optimization of entity embeddings for ECFP6 categorical variables allows FFN models to perform feature engineering in a data driven manner (Guo and Berkhahn, 2016 ). In addition, we compare the validity and efficiency of regression conformal predictors first in a retrospective test using ChEMBL25 data and subsequently in semi-prospective test using the recent Drug Design Data (D3R) Grand Challenge 4 (GC4) data set (Parks et al, 2019a ). The D3R issues blinded prediction challenges to the computer aided drug design (CADD) community to assess method performance in truly blinded scenarios.…”

Section: Introductionmentioning

confidence: 99%

“…The D3R issues blinded prediction challenges to the computer aided drug design (CADD) community to assess method performance in truly blinded scenarios. Although the data has been since released to the community, we use the data in the most recent GC4 (Parks et al, 2019a ) data set as a test set external to the data in ChEMBL25 for what we refer to as a semi-prospective test. These results show that the performance of both models suffer on the GC4 dataset relative to the ChEMBL25 validation set.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

An Analysis of Proteochemometric and Conformal Prediction Machine Learning Protein-Ligand Binding Affinity Models

parks

Gaieb

Amaro

2020

Front. Mol. Biosci.

Self Cite

View full text Add to dashboard Cite

Protein-ligand binding affinity is a key pharmacodynamic endpoint in drug discovery. Sole reliance on experimental design, make, and test cycles is costly and time consuming, providing an opportunity for computational methods to assist. Herein, we present results comparing random forest and feed-forward neural network proteochemometric models for their ability to predict pIC50 measurements for held out generic Bemis-Murcko scaffolds. In addition, we assess the ability of conformal prediction to provide calibrated prediction intervals in both a retrospective and semi-prospective test using the recently released Grand Challenge 4 data set as an external test set. In total, random forest and deep neural network proteochemometric models show quality retrospective performance but suffer in the semi-prospective setting. However, the conformal predictor prediction intervals prove to be well-calibrated both retrospectively and semi-prospectively showing that they can be used to guide hit discovery and lead optimization campaigns.

show abstract

Section: Introductionsupporting

confidence: 67%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

An Analysis of Proteochemometric and Conformal Prediction Machine Learning Protein-Ligand Binding Affinity Models

parks

Gaieb

Amaro

2020

Front. Mol. Biosci.

Self Cite

View full text Add to dashboard Cite

show abstract

“…In the case of small‐molecule ligands complexed with macromolecules of interest, the goal is often identification of lead compounds [2], or in later stages, structure‐guided lead optimization [32]. Experimentally determined ligand structures are also used to train and evaluate docking and virtual ligand screening algorithms [33]. Given the massive resources needed for drug development [34], an informed decision whether or not to use a specific target for further research is crucial.…”

Section: Resultsmentioning

confidence: 99%

Ligand‐centered assessment of SARS‐CoV‐2 drug target models in the Protein Data Bank

et al. 2020

View full text Add to dashboard Cite

A bright spot in the SARS‐CoV‐2 (CoV‐2) coronavirus pandemic has been the immediate mobilization of the biomedical community, working to develop treatments and vaccines for COVID‐19. Rational drug design against emerging threats depends on well‐established methodology, mainly utilizing X‐ray crystallography, to provide accurate structure models of the macromolecular drug targets and of their complexes with candidates for drug development. In the current crisis, the structural biological community has responded by presenting structure models of CoV‐2 proteins and depositing them in the Protein Data Bank (PDB), usually without time embargo and before publication. Since the structures from the first‐line research are produced in an accelerated mode, there is an elevated chance of mistakes and errors, with the ultimate risk of hindering, rather than speeding up, drug development. In the present work, we have used model‐validation metrics and examined the electron density maps for the deposited models of CoV‐2 proteins and a sample of related proteins available in the PDB as of April 1, 2020. We present these results with the aim of helping the biomedical community establish a better‐validated pool of data. The proteins are divided into groups according to their structure and function. In most cases, no major corrections were necessary. However, in several cases significant revisions in the functionally sensitive area of protein–inhibitor complexes or for bound ions justified correction, re‐refinement, and eventually reversioning in the PDB. The re‐refined coordinate files and a tool for facilitating model comparisons are available at https://covid-19.bioreproducibility.org. Database Validated models of CoV‐2 proteins are available in a dedicated, publicly accessible web service https://covid-19.bioreproducibility.org

show abstract

“…Although all these methods are routinely applied in drug discovery projects, improvements in their accuracy are still needed to increase their impact. [3][4][5][6] The binding free energy, ∆G • , is defined as the difference between the free energy of the ligand in solution at a standard reference concentration, and the free energy of the ligand bound to the protein. This property can be calculated in a variety of ways, but two -which we consider here -are becoming relatively more common.…”

Section: Introductionmentioning

confidence: 99%

Challenges Encountered Applying Equilibrium and Non-Equilibrium Binding Free Energy Calculations

Baumann¹,

Gapsys²,

Groot³

et al. 2020

Preprint

View full text Add to dashboard Cite

<div>Binding free energy calculations have become increasingly valuable to drive decision making in drug discovery projects. </div><div>However, among other issues, inadequate sampling can reduce accuracy, limiting the value of the technique.</div><div>In this paper we apply absolute binding free energy calculations to ligands binding to T4 lysozyme L99A and HSP90 using equilibrium and non-equilibrium approaches. We highlight sampling problems encountered in these systems, such as slow side chain rearrangements and slow changes of water placement upon ligand binding. These same types of challenges are likely to show up in other protein-ligand systems as well and we propose some strategies to diagnose and test for such problems in alchemical free energy calculations. We also explore similarities and differences in how the equilibrium and the non-equilibrium approaches handle these problems. Our results show the large amount of work still to be done to make free energy calculations robust and reliable and provide insight for future research in this area. </div>

show abstract

D3R Grand Challenge 4: Blind Prediction of Protein-Ligand Poses, Affinity Rankings, and Relative Binding Free Energies

Cited by 15 publications

References 7 publications

An Analysis of Proteochemometric and Conformal Prediction Machine Learning Protein-Ligand Binding Affinity Models

An Analysis of Proteochemometric and Conformal Prediction Machine Learning Protein-Ligand Binding Affinity Models

Ligand‐centered assessment of SARS‐CoV‐2 drug target models in the Protein Data Bank

Challenges Encountered Applying Equilibrium and Non-Equilibrium Binding Free Energy Calculations

Contact Info

Product

Resources

About