RESUMO
iQSPR is an inverse molecular design algorithm based on Bayesian inference that was developed in our previous study. Here, the algorithm is integrated in Python as a new module called iQSPR-X in the all-in-one materials informatics platform XenonPy. Our new software provides a flexible, easy-to-use, and extensible platform for users to build customized molecular design algorithms using pre-set modules and a pre-trained model library in XenonPy. In this paper, we describe key features of iQSPR-X and provide guidance on its use, illustrated by an application to a polymer design that targets a specific range of bandgap and dielectric constant.
Assuntos
Algoritmos , Teorema de Bayes , Software , Polímeros/síntese química , Polímeros/químicaRESUMO
There is a growing demand for the use of machine learning (ML) to derive fast-to-evaluate surrogate models of materials properties. In recent years, a broad array of materials property databases have emerged as part of a digital transformation of materials science. However, recent technological advances in ML are not fully exploited because of the insufficient volume and diversity of materials data. An ML framework called "transfer learning" has considerable potential to overcome the problem of limited amounts of materials data. Transfer learning relies on the concept that various property types, such as physical, chemical, electronic, thermodynamic, and mechanical properties, are physically interrelated. For a given target property to be predicted from a limited supply of training data, models of related proxy properties are pretrained using sufficient data; these models capture common features relevant to the target task. Repurposing of such machine-acquired features on the target task yields outstanding prediction performance even with exceedingly small data sets, as if highly experienced human experts can make rational inferences even for considerably less experienced tasks. In this study, to facilitate widespread use of transfer learning, we develop a pretrained model library called XenonPy.MDL. In this first release, the library comprises more than 140â¯000 pretrained models for various properties of small molecules, polymers, and inorganic crystalline materials. Along with these pretrained models, we describe some outstanding successes of transfer learning in different scenarios such as building models with only dozens of materials data, increasing the ability of extrapolative prediction through a strategic model transfer, and so on. Remarkably, transfer learning has autonomously identified rather nontrivial transferability across different properties transcending the different disciplines of materials science; for example, our analysis has revealed underlying bridges between small molecules and polymers and between organic and inorganic chemistry.
RESUMO
Peptides with cell attachment activity are beneficial component of biomaterials for tissue engineering. Conformational structure is one of the important factors for the biological activities. The EF1 peptide (DYATLQLQEGRLHFMFDLG) derived from laminin promotes cell spreading and cell attachment activity mediated by α2ß1 integrin. Although the sequence of the EF2 peptide (DFATVQLRNGFPYFSYDLG) is homologous sequence to that of EF1, EF2 does not promote cell attachment activity. To determine whether there are structural differences between EF1 and EF2, we performed replica exchange molecular dynamics (REMD) simulations and conventional molecular dynamics (MD) simulations. We found that EF1 and EF2 had ß-sheet structure as a secondary structure around the global minimum. However, EF2 had variety of structures around the global minimum compared with EF1 and has easily escaped from the bottom of free energy. The structural fluctuation of the EF1 is smaller than that of the EF2. The structural variation of EF2 is related to these differences in the structural fluctuation and the number of the hydrogen bonds (H-bonds). From the analysis of H-bonds in the ß-sheet, the number of H-bonds in EF1 is larger than that in EF2 in the time scale of the conventional MD simulation, suggesting that the formation of H-bonds is related to the differences in the structural fluctuation between EF1 and EF2. From the analysis of other non-covalent interactions in the amino acid sequences of EF1 and EF2, EF1 has three pairs of residues with hydrophobic interaction, and EF2 has two pairs. These results indicate that several non-covalent interactions are important for structural stabilization. Consequently, the structure of EF1 is stabilized by H-bonds and pairs of hydrophobic amino acids in the terminals. Hence, we propose that non-covalent interactions around N-terminal and C-terminal of the peptides are crucial for maintaining the ß-sheet structure of the peptides.