Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
162 tokens/sec
GPT-4o
7 tokens/sec
Gemini 2.5 Pro Pro
45 tokens/sec
o3 Pro
4 tokens/sec
GPT-4.1 Pro
38 tokens/sec
DeepSeek R1 via Azure Pro
28 tokens/sec
2000 character limit reached

Towards End-to-End Structure Solutions from Information-Compromised Diffraction Data via Generative Deep Learning (2312.15136v1)

Published 23 Dec 2023 in physics.comp-ph, cs.AI, and cs.CV

Abstract: The revolution in materials in the past century was built on a knowledge of the atomic arrangements and the structure-property relationship. The sine qua non for obtaining quantitative structural information is single crystal crystallography. However, increasingly we need to solve structures in cases where the information content in our input signal is significantly degraded, for example, due to orientational averaging of grains, finite size effects due to nanostructure, and mixed signals due to sample heterogeneity. Understanding the structure property relationships in such situations is, if anything, more important and insightful, yet we do not have robust approaches for accomplishing it. In principle, ML and deep learning (DL) are promising approaches since they augment information in the degraded input signal with prior knowledge learned from large databases of already known structures. Here we present a novel ML approach, a variational query-based multi-branch deep neural network that has the promise to be a robust but general tool to address this problem end-to-end. We demonstrate the approach on computed powder x-ray diffraction (PXRD), along with partial chemical composition information, as input. We choose as a structural representation a modified electron density we call the Cartesian mapped electron density (CMED), that straightforwardly allows our ML model to learn material structures across different chemistries, symmetries and crystal systems. When evaluated on theoretically simulated data for the cubic and trigonal crystal systems, the system achieves up to $93.4\%$ average similarity with the ground truth on unseen materials, both with known and partially-known chemical composition information, showing great promise for successful structure solution even from degraded and incomplete input data.

Definition Search Book Streamline Icon: https://streamlinehq.com
References (64)
  1. Giacovazzo, C. Fundamentals of crystallography, vol. 7 (Oxford university press, USA, 2002).
  2. Hammond, C. The basics of crystallography and diffraction, vol. 21 (International Union of Crystallography texts on crystallography, 2015).
  3. The crystal structure of the alums. \JournalTitleProceedings of the Royal Society of London. Series A-Mathematical and Physical Sciences 148, 664–680 (1935).
  4. Powder diffraction: theory and practice (Royal society of chemistry, 2008).
  5. An x-ray study of the dissociation of an alloy of copper, iron and nickel. \JournalTitleProceedings of the Royal Society of London. Series A. Mathematical and Physical Sciences 181, 368–378 (1943).
  6. The structure of graphite. \JournalTitleProceedings of the Royal Society of London. Series A. Mathematical and Physical Sciences 181, 101–105 (1942).
  7. Lipson, H. The study of metals and alloys by X-ray powder diffraction methods (University College Cardiff Press Cardiff, 1984).
  8. The problem with determining atomic structure at the nanoscale. \JournalTitlescience 316, 561–565 (2007).
  9. Structure determination from powder diffraction data. \JournalTitleActa Crystallogr A Found Crystallogr 64, 52–64, DOI: 10.1107/S0108767307064252 (2008).
  10. Jumper, J. et al. Highly accurate protein structure prediction with alphafold. \JournalTitleNature 596, 583–589 (2021).
  11. Improved prediction of protein-protein interactions using alphafold2. \JournalTitleNature communications 13, 1265 (2022).
  12. Baek, M. et al. Accurate prediction of protein structures and interactions using a three-track neural network. \JournalTitleScience 373, 871–876 (2021).
  13. Dobson, C. M. Protein folding and misfolding. \JournalTitleNature 426, 884–890 (2003).
  14. Imagenet classification with deep convolutional neural networks. \JournalTitleAdvances in neural information processing systems 25 (2012).
  15. Bojarski, M. et al. End to end learning for self-driving cars. \JournalTitlearXiv preprint arXiv:1604.07316 (2016).
  16. Amodei, D. et al. Deep speech 2: End-to-end speech recognition in english and mandarin. In International conference on machine learning, 173–182 (PMLR, 2016).
  17. Using a machine learning approach to determine the space group of a structure from the atomic pair distribution function. \JournalTitleActa Cryst A 75, 633–643, DOI: 10.1107/S2053273319005606 (2019).
  18. Oviedo, F. et al. Fast and interpretable classification of small x-ray diffraction datasets using data augmentation and deep neural networks. \JournalTitlenpj Computational Materials 5, 60 (2019).
  19. Suzuki, Y. et al. Symmetry prediction and knowledge discovery from x-ray diffraction patterns using an interpretable machine learning approach. \JournalTitleScientific reports 10, 21790 (2020).
  20. Park, W. B. et al. Classification of crystal structure using a convolutional neural network. \JournalTitleIUCrJ 4, 486–494 (2017).
  21. A deep-learning technique for phase identification in multiphase inorganic compounds using synthetic xrd powder patterns. \JournalTitleNature communications 11, 86 (2020).
  22. Decoding crystallography from high-resolution electron imaging and diffraction datasets with deep learning. \JournalTitleScience Advances 5, eaaw1949 (2019).
  23. Insightful classification of crystal structures using deep learning. \JournalTitleNature communications 9, 2775 (2018).
  24. Identification of crystal symmetry from noisy diffraction patterns by a shape analysis and deep learning. \JournalTitlenpj Computational Materials 6, 196 (2020).
  25. Garcia-Cardona, C. et al. Learning to predict material structure from neutron scattering data. In 2019 IEEE International Conference on Big Data (Big Data), 4490–4497 (IEEE, 2019).
  26. Merker, H. A. et al. Machine learning magnetism classifiers from atomic coordinates. \JournalTitleIscience 25 (2022).
  27. Merchant, A. et al. Scaling deep learning for materials discovery. \JournalTitleNature 1–6 (2023).
  28. Yang, M. et al. Scalable diffusion for materials discovery. \JournalTitlearXiv e-prints (2023).
  29. Hernández-García, A. et al. Crystal-gflownet: sampling materials with desirable properties and constraints. In AI for Accelerated Materials Design-NeurIPS 2023 Workshop (2023).
  30. A deep learning solution for crystallographic structure determination. \JournalTitleIUCrJ 10, 487–496 (2023).
  31. Dun, C. et al. Crysformer: Protein structure prediction via 3d patterson maps and partial structure attention. \JournalTitlearXiv preprint arXiv:2310.03899 (2023).
  32. The x-ray crystallography phase problem solved thanks to alphafold and rosettafold models: a case-study report. \JournalTitleActa Crystallographica Section D: Structural Biology 78, 517–531 (2022).
  33. Kjær, E. T. S. et al. DeepStruc: Towards structure solution from pair distribution function data using deep generative models. \JournalTitleDigital Discovery 2, 69–80, DOI: 10.1039/D2DD00086E (2023).
  34. Auto-encoding variational bayes. \JournalTitlearXiv preprint arXiv:1312.6114 (2013).
  35. pixelnerf: Neural radiance fields from one or few images. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 4578–4587 (2021).
  36. Mildenhall, B. et al. Nerf: Representing scenes as neural radiance fields for view synthesis. \JournalTitleCommunications of the ACM 65, 99–106 (2021).
  37. Tancik, M. et al. Fourier features let networks learn high frequency functions in low dimensional domains. \JournalTitleAdvances in Neural Information Processing Systems 33, 7537–7547 (2020).
  38. Scene representation networks: Continuous 3d-structure-aware neural scene representations. \JournalTitleAdvances in Neural Information Processing Systems 32 (2019).
  39. Deepsdf: Learning continuous signed distance functions for shape representation. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 165–174 (2019).
  40. Hoffmann, J. et al. Data-Driven Approach to Encoding and Decoding 3-D Crystal Structures (2019). 1909.00949.
  41. Higgins, I. et al. beta-vae: Learning basic visual concepts with a constrained variational framework. In International conference on learning representations (2016).
  42. Jain, A. et al. Commentary: The materials project: A materials genome approach to accelerating materials innovation. \JournalTitleAPL materials 1 (2013).
  43. Image quality assessment: from error visibility to structural similarity. \JournalTitleIEEE transactions on image processing 13, 600–612 (2004).
  44. Image quality metrics: Psnr vs. ssim. In 2010 20th international conference on pattern recognition, 2366–2369 (IEEE, 2010).
  45. New developments in the inorganic crystal structure database (icsd): accessibility in support of materials research and design. \JournalTitleActa Crystallographica Section B: Structural Science 58, 364–369 (2002).
  46. Hafner, J. Ab-initio simulations of materials using vasp: Density-functional theory and beyond. \JournalTitleJournal of computational chemistry 29, 2044–2078 (2008).
  47. The materials project workshop. https://workshop.materialsproject.org/lessons/01_website_walkthrough/website_walkthrough/. Retrieved 08/25/2023.
  48. Structure of materials: an introduction to crystallography, diffraction and symmetry (Cambridge University Press, 2012).
  49. Diffraction patterns: How diffraction patterns are calculated on the materials project (mp) website. https://docs.materialsproject.org/methodology/materials-methodology/diffraction-patterns. Retrieved 08/25/2023.
  50. Batch normalization: Accelerating deep network training by reducing internal covariate shift. In International conference on machine learning, 448–456 (pmlr, 2015).
  51. Charge density: Obtaining the charge density shown on the materials project (mp) website. https://docs.materialsproject.org/methodology/materials-methodology/charge-density. Retrieved 08/25/2023.
  52. Chgcar. https://www.vasp.at/wiki/index.php/CHGCAR. Retrieved 08/25/2023.
  53. Shen, J.-X. et al. A representation-independent electronic charge density database for crystalline materials. \JournalTitleScientific Data 9, 661 (2022).
  54. mp-pyrho. https://github.com/materialsproject/pyrho. Retrieved 08/25/2023.
  55. Chitturi, S. R. et al. Automated prediction of lattice parameters from x-ray powder diffraction patterns. \JournalTitleJournal of Applied Crystallography 54, 1799–1810 (2021).
  56. Towards the extraction of the crystal cell parameters from pair distribution function profiles. \JournalTitleIUCrJ 10, 610–623, DOI: 10.1107/S2052252523006887 (2023).
  57. Densely connected convolutional networks. In Proceedings of the IEEE conference on computer vision and pattern recognition, 4700–4708 (2017).
  58. Layer normalization. \JournalTitlearXiv preprint arXiv:1607.06450 (2016).
  59. Neural tangent kernel: Convergence and generalization in neural networks. \JournalTitleAdvances in neural information processing systems 31 (2018).
  60. Film: Visual reasoning with a general conditioning layer. In Proceedings of the AAAI conference on artificial intelligence, vol. 32 (2018).
  61. On information and sufficiency. \JournalTitleThe annals of mathematical statistics 22, 79–86 (1951).
  62. Adam: A method for stochastic optimization. \JournalTitlearXiv preprint arXiv:1412.6980 (2014).
  63. Sgdr: Stochastic gradient descent with warm restarts. \JournalTitlearXiv preprint arXiv:1608.03983 (2016).
  64. Van der Walt, S. et al. scikit-image: image processing in python. \JournalTitlePeerJ 2, e453 (2014).
Citations (3)
View on Semantic Scholar

Summary

We haven't generated a summary for this paper yet.

Ai Sparkles Streamline Icon: https://streamlinehq.com Summarize Now