-
RapidDock: Unlocking Proteome-scale Molecular Docking
Authors:
Rafał Powalski,
Bazyli Klockiewicz,
Maciej Jaśkowski,
Bartosz Topolski,
Paweł Dąbrowski-Tumański,
Maciej Wiśniewski,
Łukasz Kuciński,
Piotr Miłoś,
Dariusz Plewczynski
Abstract:
Accelerating molecular docking -- the process of predicting how molecules bind to protein targets -- could boost small-molecule drug discovery and revolutionize medicine. Unfortunately, current molecular docking tools are too slow to screen potential drugs against all relevant proteins, which often results in missed drug candidates or unexpected side effects occurring in clinical trials. To addres…
▽ More
Accelerating molecular docking -- the process of predicting how molecules bind to protein targets -- could boost small-molecule drug discovery and revolutionize medicine. Unfortunately, current molecular docking tools are too slow to screen potential drugs against all relevant proteins, which often results in missed drug candidates or unexpected side effects occurring in clinical trials. To address this gap, we introduce RapidDock, an efficient transformer-based model for blind molecular docking. RapidDock achieves at least a $100 \times$ speed advantage over existing methods without compromising accuracy. On the Posebusters and DockGen benchmarks, our method achieves $52.1\%$ and $44.0\%$ success rates ($\text{RMSD}<2$Å), respectively. The average inference time is $0.04$ seconds on a single GPU, highlighting RapidDock's potential for large-scale docking studies. We examine the key features of RapidDock that enable leveraging the transformer architecture for molecular docking, including the use of relative distance embeddings of $3$D structures in attention matrices, pre-training on protein folding, and a custom loss function invariant to molecular symmetries.
△ Less
Submitted 16 October, 2024;
originally announced November 2024.
-
Role of the host genetic variability in the influenza-A virus susceptibility; a review
Authors:
Ana Carolina Arcanjo,
Giovanni Mazzocco,
Silviene Fabiana de Oliveira,
Dariusz Plewczynski,
Jan P. Radomski
Abstract:
The aftermath of influenza infection is determined by a complex set of host-pathogen interactions, where genomic variability on both viral and host sides influences the final outcome. Although there exists large body of literature describing influenza virus variability, only a very small fraction covers the issue of host variance. The goal of this review is to explore the variability of host genes…
▽ More
The aftermath of influenza infection is determined by a complex set of host-pathogen interactions, where genomic variability on both viral and host sides influences the final outcome. Although there exists large body of literature describing influenza virus variability, only a very small fraction covers the issue of host variance. The goal of this review is to explore the variability of host genes responsible for host-pathogen interactions, paying particular attention to genes responsible for the presence of sialylated glycans in the host endothelial membrane, mucus, genes used by viral immune escape mechanisms, and genes particularly expressed after vaccination, since they are more likely to have a direct influence on the infection outcome.
△ Less
Submitted 6 September, 2014;
originally announced September 2014.
-
Landau Theory of Adaptive Integration in Computational Intelligence
Authors:
Dariusz Plewczynski
Abstract:
Computational Intelligence (CI) is a sub-branch of Artificial Intelligence paradigm focusing on the study of adaptive mechanisms to enable or facilitate intelligent behavior in complex and changing environments. There are several paradigms of CI [like artificial neural networks, evolutionary computations, swarm intelligence, artificial immune systems, fuzzy systems and many others], each of these…
▽ More
Computational Intelligence (CI) is a sub-branch of Artificial Intelligence paradigm focusing on the study of adaptive mechanisms to enable or facilitate intelligent behavior in complex and changing environments. There are several paradigms of CI [like artificial neural networks, evolutionary computations, swarm intelligence, artificial immune systems, fuzzy systems and many others], each of these has its origins in biological systems [biological neural systems, natural Darwinian evolution, social behavior, immune system, interactions of organisms with their environment]. Most of those paradigms evolved into separate machine learning (ML) techniques, where probabilistic methods are used complementary with CI techniques in order to effectively combine elements of learning, adaptation, evolution and Fuzzy logic to create heuristic algorithms that are, in some sense, intelligent. The current trend is to develop consensus techniques, since no single machine learning algorithms is superior to others in all possible situations. In order to overcome this problem several meta-approaches were proposed in ML focusing on the integration of results from different methods into single prediction. We discuss here the Landau theory for the nonlinear equation that can describe the adaptive integration of information acquired from an ensemble of independent learning agents. The influence of each individual agent on other learners is described similarly to the social impact theory. The final decision outcome for the consensus system is calculated using majority rule in the stationary limit, yet the minority solutions can survive inside the majority population as the complex intermittent clusters of opposite opinion.
△ Less
Submitted 9 June, 2010;
originally announced June 2010.
