Antibody–like Synthetic Molecular Recognition Thin Layers Fabricated by Molecular Imprinting Based on Specific Protein–ligand Interactions

MEMBRANE ◽  
2017 ◽  
Vol 42 (3) ◽  
pp. 97-103
Author(s):  
Yuri Kamon ◽  
Toshifumi Takeuchi
Keyword(s):  
Molecular Recognition ◽  
Molecular Imprinting ◽  
Specific Protein ◽  
Thin Layers ◽  
Protein Ligand Interactions ◽  
Ligand Interactions

scholarly journals Hidden Bias in the DUD-E Dataset Leads to Misleading Performance of Deep Learning in Structure-Based Virtual Screening

2019 ◽  
Author(s):  
Lieyang Chen ◽  
Anthony Cruz ◽  
Steven Ramsey ◽  
Callum J. Dickson ◽  
José S. Duca ◽  
...  
Keyword(s):  
Deep Learning ◽  
Virtual Screening ◽  
Molecular Recognition ◽  
Autodock Vina ◽  
Methodology Development ◽  
X Ray ◽  
Learning Biases ◽  
Protein Ligand Interactions ◽  
Ligand Interactions ◽  
Docking Program

<p>Recently much effort has been invested in using convolutional neural network (CNN) models trained on 3D structural images of protein-ligand complexes to distinguish binding from non-binding ligands for virtual screening. However, the dearth of reliable protein-ligand x-ray structures and binding affinity data has required the use of constructed datasets for the training and evaluation of CNN molecular recognition models. Here, we outline various sources of bias in one such widely-used dataset, the Directory of Useful Decoys: Enhanced (DUD-E). We have constructed and performed tests to investigate whether CNN models developed using DUD-E are properly learning the underlying physics of molecular recognition, as intended, or are instead learning biases inherent in the dataset itself. We find that superior enrichment efficiency in CNN models can be attributed to the analogue and decoy bias hidden in the DUD-E dataset rather than successful generalization of the pattern of protein-ligand interactions. Comparing additional deep learning models trained on PDBbind datasets, we found that their enrichment performances using DUD-E are not superior to the performance of the docking program AutoDock Vina. Together, these results suggest that biases that could be present in constructed datasets should be thoroughly evaluated before applying them to machine learning based methodology development. </p>

Molecular recognition by proteins: Protein-ligand interactions from a structural perspective

1996 ◽  
Vol 24 (1) ◽  
pp. 280-284 ◽  
Author(s):  
A. C. Wallace ◽  
R. A. Laskowski ◽  
J. Singh ◽  
J. M. Thornton
Keyword(s):  
Molecular Recognition ◽  
Protein Ligand Interactions ◽  
Ligand Interactions ◽  
Structural Perspective

scholarly journals Probing molecular specificity with deep sequencing and biophysically interpretable machine learning

2021 ◽  
Author(s):  
H. Tomas Rube ◽  
Chaitanya Rastogi ◽  
Siqian Feng ◽  
Judith Franziska Kribelbauer ◽  
Allyson Li ◽  
...  
Keyword(s):  
Machine Learning ◽  
Biological Networks ◽  
Specific Protein ◽  
Biophysical Parameters ◽  
Kinase Substrate ◽  
Protein Ligand Interactions ◽  
Sequence Recognition ◽  
Ligand Interactions ◽  
The Impact

Quantifying sequence-specific protein-ligand interactions is critical for understanding and exploiting numerous cellular processes, including gene regulation and signal transduction. Next-generation sequencing (NGS) based assays are increasingly being used to profile these interactions with high-throughput. However, these assays do not provide the biophysical parameters that have long been used to uncover the quantitative rules underlying sequence recognition. We developed a highly flexible machine learning framework, called ProBound, to define sequence recognition in terms of biophysical parameters based on NGS data. ProBound quantifies transcription factor (TF) behavior with models that accurately predict binding affinity over a range exceeding that of previous resources, captures the impact of DNA modifications and conformational flexibility of multi-TF complexes, and infers specificity directly from in vivo data such as ChIP-seq without peak calling. When coupled with a new assay called Kd-seq, it determines the absolute affinity of protein-ligand interactions. It can also profile the kinetics of kinase-substrate interactions. By constructing a biophysically robust foundation for profiling sequence recognition, ProBound opens up new avenues for decoding biological networks and rationally engineering protein-ligand interactions.

scholarly journals Hidden Bias in the DUD-E Dataset Leads to Misleading Performance of Deep Learning in Structure-Based Virtual Screening

2019 ◽  
Author(s):  
Lieyang Chen ◽  
Anthony Cruz ◽  
Steven Ramsey ◽  
Callum J. Dickson ◽  
José S. Duca ◽  
...  
Keyword(s):  
Deep Learning ◽  
Virtual Screening ◽  
Molecular Recognition ◽  
Autodock Vina ◽  
Methodology Development ◽  
X Ray ◽  
Learning Biases ◽  
Protein Ligand Interactions ◽  
Ligand Interactions ◽  
Docking Program

<p>Recently much effort has been invested in using convolutional neural network (CNN) models trained on 3D structural images of protein-ligand complexes to distinguish binding from non-binding ligands for virtual screening. However, the dearth of reliable protein-ligand x-ray structures and binding affinity data has required the use of constructed datasets for the training and evaluation of CNN molecular recognition models. Here, we outline various sources of bias in one such widely-used dataset, the Directory of Useful Decoys: Enhanced (DUD-E). We have constructed and performed tests to investigate whether CNN models developed using DUD-E are properly learning the underlying physics of molecular recognition, as intended, or are instead learning biases inherent in the dataset itself. We find that superior enrichment efficiency in CNN models can be attributed to the analogue and decoy bias hidden in the DUD-E dataset rather than successful generalization of the pattern of protein-ligand interactions. Comparing additional deep learning models trained on PDBbind datasets, we found that their enrichment performances using DUD-E are not superior to the performance of the docking program AutoDock Vina. Together, these results suggest that biases that could be present in constructed datasets should be thoroughly evaluated before applying them to machine learning based methodology development. </p>

scholarly journals Could Quantum Mechanical Properties be Reflected on Classical Molecular Dynamics? the Case of Halogenated Organic Compounds of Biological Interest

2019 ◽  
Author(s):  
Lucas dos Santos Azevedo Azevedo ◽  
Ingrid G. Prandi ◽  
Teodorico Ramalho
Keyword(s):  
Drug Design ◽  
Force Fields ◽  
Noncovalent Interactions ◽  
Computational Cost ◽  
Specific Protein ◽  
Electronic Effects ◽  
Protein Ligand Interactions ◽  
Ligand Interactions ◽  
Biological Target ◽  
Hole Model

<p>Essential to understanding life, the biomolecular phenomena have been an important subject in science, therefore a necessary path to be covered to make progress in human knowledge. To fully comprehend these processes, the noncovalent interactions are the key. In this review, we discuss how specific protein-ligand interactions can be efficiently described by low computational cost methods, such as Molecular Mechanics (MM). We have taken as example the case of the halogen bonds (XB). Albeit generally weaker than the hydrogen bonds (HB), the XBs play a key role to drug design, enhancing the affinity and selectivity towards the biological target. Along with the attraction between two electronegative atoms in XBs explained by the σ-hole model, important orbital interactions, as well as relief of Pauli repulsion take place. Nonetheless, such electronic effects can be only well described by accurate quantum chemical methods that have strong limitations dealing with supramolecular systems due to their high computational cost. To go beyond the poor description of XBs by MM methods, reparametrizing the force-fields equations can be a way to keep the balance between accuracy and computational cost. Thus, we have shown the steps to be considered when parametrizing force-fields to achieve reliable results of complex noncovalent interactions at MM level for In Silico drug design methods.</p>

Sign in / Sign up

Export Citation Format

Share Document