scholarly journals Targeted sequence design within the coarse-grained polymer genome

2020 ◽  
Vol 6 (43) ◽  
pp. eabc6216
Author(s):  
Michael A. Webb ◽  
Nicholas E. Jackson ◽  
Phwey S. Gil ◽  
Juan J. de Pablo
Keyword(s):  
Machine Learning ◽  
Materials Science ◽  
Chain Structure ◽  
Data Availability ◽  
Data Driven ◽  
Coarse Grained ◽  
Grand Challenge ◽  
Sequence Information ◽  
Single Chain ◽  
Sequence Design

The chemical design of polymers with target structural and/or functional properties represents a grand challenge in materials science. While data-driven design approaches are promising, success with polymers has been limited, largely due to limitations in data availability. Here, we demonstrate the targeted sequence design of single-chain structure in polymers by combining coarse-grained modeling, machine learning, and model optimization. Nearly 2000 unique coarse-grained polymers are simulated to construct and analyze machine learning models. We find that deep neural networks inexpensively and reliably predict structural properties with limited sequence information as input. By coupling trained ML models with sequential model-based optimization, polymer sequences are proposed to exhibit globular, swollen, or rod-like behaviors, which are verified by explicit simulations. This work highlights the promising integration of coarse-grained modeling with data-driven design and represents a necessary and crucial step toward more complex polymer design efforts.

Evolving the Materials Genome: How Machine Learning Is Fueling the Next Generation of Materials Discovery

2020 ◽  
Vol 50 (1) ◽  
pp. 1-25 ◽  
Author(s):  
Changwon Suh ◽  
Clyde Fare ◽  
James A. Warren ◽  
Edward O. Pyzer-Knapp
Keyword(s):  
Machine Learning ◽  
Materials Science ◽  
State Of The Art ◽  
Learning Technologies ◽  
Machine Learning Algorithms ◽  
Data Driven ◽  
Computational Learning ◽  
Data Intensive ◽  
Research Activities ◽  
Materials Genome

Machine learning, applied to chemical and materials data, is transforming the field of materials discovery and design, yet significant work is still required to fully take advantage of machine learning algorithms, tools, and methods. Here, we review the accomplishments to date of the community and assess the maturity of state-of-the-art, data-intensive research activities that combine perspectives from materials science and chemistry. We focus on three major themes—learning to see, learning to estimate, and learning to search materials—to show how advanced computational learning technologies are rapidly and successfully used to solve materials and chemistry problems. Additionally, we discuss a clear path toward a future where data-driven approaches to materials discovery and design are standard practice.

Multiscale modeling of polymers closely coupled to Broad Q neutron scattering from NIMROD

2013 ◽  
Vol 1524 ◽  
Author(s):  
Thomas Gkourmpis ◽  
Daniel Lopez ◽  
Geoffrey R. Mitchell
Keyword(s):  
Structure Factor ◽  
Chain Structure ◽  
Distribution Functions ◽  
Chain Conformation ◽  
Coarse Grained ◽  
Lamellar Thickness ◽  
Single Chain ◽  
Crystal Unit Cell ◽  
Key Length ◽  
Using Data

ABSTRACTWe use data over an extended Q range from 0.01 to 100Å-1 from the recently commissioned NIMROD instrument at the ISIS pulsed neutron source to develop a multi-scale inverse modeling procedure which will provide insight in to the phase transformations of polymer systems. The first level of our procedure is atomistic and we use internal coordinates (bond length, bond angles and torsion angles) to define the polymer chain in full atomistic detail. Values were assigned to each internal coordinate within the chain using a stochastic Monte Carlo method in which the probabilities were drawn from distributions representing the possible range of values. Using this approach, random chain configurations could be rapidly built and the intrachain structure factor calculated utilizing a small set of parameters and compared with the experimental function. Parameters representing the probability distribution functions were systematically varied using a grid search to find the values which gave the best fit to the structure factor for Q > 3Å-1 in order to determine the details of the chain conformation in the molten phase. This process was repeated for data over the same extended Q range obtained at lower temperatures where the polymer was expected to crystallize. Polymers crystallize via chainfolded thin lamellae crystals. Such crystals give rise to an intense peak at Q ∼ 0.03Å-1. This scattering can be calculated using a lamellar stack model, coarse-grained from the single chain structure. We describe this approach using data obtained on the crystallization from the melt phase of perdeuterated polymers. The objective here is to follow the three key length scales; the chain folded lamellar thickness of ∼ 10nm, the crystal unit cell ∼ 1nm and the detail of the chain conformation is ∼ 0.1nm.

Machine learning for data-driven discovery in solid Earth geoscience

Science ◽  
2019 ◽  
Vol 363 (6433) ◽  
pp. eaau0323 ◽  
Author(s):  
Karianne J. Bergen ◽  
Paul A. Johnson ◽  
Maarten V. de Hoop ◽  
Gregory C. Beroza
Keyword(s):  
Machine Learning ◽  
Solid Earth ◽  
Computer Simulations ◽  
Direct Observation ◽  
The State ◽  
Important Task ◽  
Data Availability ◽  
Data Driven ◽  
State Of The Field

Understanding the behavior of Earth through the diverse fields of the solid Earth geosciences is an increasingly important task. It is made challenging by the complex, interacting, and multiscale processes needed to understand Earth’s behavior and by the inaccessibility of nearly all of Earth’s subsurface to direct observation. Substantial increases in data availability and in the increasingly realistic character of computer simulations hold promise for accelerating progress, but developing a deeper understanding based on these capabilities is itself challenging. Machine learning will play a key role in this effort. We review the state of the field and make recommendations for how progress might be broadened and accelerated.

scholarly journals Accurate model of liquid-liquid phase behaviour of intrinsically-disordered proteins from data-driven optimization of single-chain properties

2021 ◽  
Author(s):  
Giulio Tesei ◽  
Thea K. Schulze ◽  
Ramon Crehuet ◽  
Kresten Lindorff-Larsen
Keyword(s):  
Liquid Phase ◽  
Intrinsically Disordered Proteins ◽  
Molecular Simulations ◽  
Phase Behaviour ◽  
Data Driven ◽  
Coarse Grained ◽  
Disordered Proteins ◽  
Model Parameters ◽  
Single Chain ◽  
Intrinsically Disordered

Many intrinsically disordered proteins (IDPs) may undergo liquid-liquid phase separation (LLPS) and participate in the formation of membraneless organelles in the cell, thereby contributing to the regulation and compartmentalisation of intracellular biochemical reactions. The phase behaviour of IDPs is sequence-dependent, and its investigation through molecular simulations requires protein models that combine computational efficiency with an accurate description of intra- and intermolecular interactions. We developed a general coarse-grained model of IDPs, with residue-level detail, based on an extensive set of experimental data on single-chain properties. Ensemble-averaged experimental observables are predicted from molecular simulations, and a data-driven parameter-learning procedure is used to identify the residue-specific model parameters that minimize the discrepancy between predictions and experiments. The model accurately reproduces the experimentally observed conformational propensities of a set of IDPs. Through two-body as well as large-scale molecular simulations, we show that the optimization of the intramolecular interactions results in improved predictions of protein self-association and LLPS.

Rapid Isolation of High Molecular Weight Urokinase from Native Human Urine

1982 ◽  
Vol 47 (03) ◽  
pp. 197-202 ◽  
Author(s):  
Kurt Huber ◽  
Johannes Kirchheimer ◽  
Bernd R Binder
Keyword(s):  
Molecular Weight ◽  
High Molecular Weight ◽  
Human Urine ◽  
Gel Filtration ◽  
Chain Structure ◽  
The Other ◽  
Single Chain ◽  
Apparent Homogeneity ◽  
Sequential Adsorption

SummaryUrokinase (UK) could be purified to apparent homogeneity starting from crude urine by sequential adsorption and elution of the enzyme to gelatine-Sepharose and agmatine-Sepharose followed by gel filtration on Sephadex G-150. The purified product exhibited characteristics of the high molecular weight urokinase (HMW-UK) but did contain two distinct entities, one of which exhibited a two chain structure as reported for the HMW-UK while the other one exhibited an apparent single chain structure. The purification described is rapid and simple and results in an enzyme with probably no major alterations. Yields are high enough to obtain purified enzymes for characterization of UK from individual donors.

Evolution of Metastable Structures in Bimetallic Catalysts from Microscopy and Machine-Learning Molecular Dynamics

2020 ◽  
Author(s):  
Jin Soo Lim ◽  
Jonathan Vandermause ◽  
Matthijs A. van Spronsen ◽  
Albert Musaelian ◽  
Christopher R. O’Connor ◽  
...  
Keyword(s):  
Machine Learning ◽  
Molecular Dynamics ◽  
Large Scale ◽  
Materials Science ◽  
Complete Characterization ◽  
Layer By Layer ◽  
Surface Restructuring ◽  
Metastable Structures ◽  
Mechanistic Investigation ◽  
Underlying Mechanisms

Restructuring of interface plays a crucial role in materials science and heterogeneous catalysis. Bimetallic systems, in particular, often adopt very different composition and morphology at surfaces compared to the bulk. For the first time, we reveal a detailed atomistic picture of the long-timescale restructuring of Pd deposited on Ag, using microscopy, spectroscopy, and novel simulation methods. Encapsulation of Pd by Ag always precedes layer-by-layer dissolution of Pd, resulting in significant Ag migration out of the surface and extensive vacancy pits. These metastable structures are of vital catalytic importance, as Ag-encapsulated Pd remains much more accessible to reactants than bulk-dissolved Pd. The underlying mechanisms are uncovered by performing fast and large-scale machine-learning molecular dynamics, followed by our newly developed method for complete characterization of atomic surface restructuring events. Our approach is broadly applicable to other multimetallic systems of interest and enables the previously impractical mechanistic investigation of restructuring dynamics.

Sign in / Sign up

Export Citation Format

Share Document