Toms Rekis
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Research Interests
Toms Rekis has a total of 104 co-authors. Here are the top co-authors by number of shared publications:
- Agris Bērziņš
Department of Chemistry, Faculty of Medicine and Life Sciences University of Latvia Riga Latvia
14 shared publications - Sander van Smaalen
University of Bayreuth
12 shared publications Andris Actiņš
Faculty of Chemistry, University of Latvia, Jelgavas iela 1, Riga, LV-1004, Latvia
8 shared publications- Andreas Schönleber
Laboratory of Crystallography, University of Bayreuth, 95447 Bayreuth, Germany
7 shared publications Carsten Paulmann
P24, PETRA III, Deutsches Elektronen-Synchrotron DESY, Notkestrasse 85, 22607 Hamburg, Germany
7 shared publicationsMartin Tolkiehn
P24, PETRA III, Deutsches Elektronen-Synchrotron DESY, Notkestrasse 85, 22607 Hamburg, Germany
7 shared publications
Latest Publications
- 1 January 2027
- Published by International Union of Crystallography (IUCr) in Acta Crystallographica Section A Foundations and Advances
- Vol. 83 (1)
Towards quantum crystallographic description of incommensurately modulated structures
In this contribution, a slight modification of the structure-factor equation for incommensurately modulated structures is proposed in order to enable structure description with non-spherical atomic form factors. A hypothetical (2 + 1)-dimensional structure in Flatland is used as an example to demonstrate the need for the modification when switching to a more advanced description of the electron density. The study is primarily concerned with molecular systems.
- 1 January 2027
- Published by International Union of Crystallography (IUCr) in Acta Crystallographica Section A Foundations and Advances
- Vol. 83 (1)
Towards quantum crystallographic description of incommensurately modulated structures
In this contribution, a slight modification of the structure-factor equation for incommensurately modulated structures is proposed in order to enable structure description with non-spherical atomic form factors. A hypothetical (2 + 1)-dimensional structure in Flatland is used as an example to demonstrate the need for the modification when switching to a more advanced description of the electron density. The study is primarily concerned with molecular systems.
- Open Access
- 6 October 2026
- Published by Springer Nature in Communications Chemistry
- Vol. 9 (1) , 314
Pair distribution function “feat.” solid-state NMR in a combined approach for structure determination of organic molecules
Investigating the local structure of molecular solids is critical for establishing structure–property relationships and controlling material performance. Pair distribution function (PDF) analysis probes poorly crystalline, nanocrystalline and amorphous materials, yet its application to organic compounds remains limited. The recently developed Global-PDF-Fit method enables ab initio crystal structure determination of organic molecules, although intrinsic limitations can complicate unique structure identification. Here, we present a synergistic workflow integrating Global-PDF-Fit with solid-state NMR (SSNMR), demonstrating how restraints and intermolecular proximity constraints from 1D magic angle spinning (MAS) and 2D 1 H- 13 C double cross polarization and 1 H double quantum/ 1 H single quantum MAS SSNMR experiments reduce computational time, guide structure ranking, and increase reliability, enabling unambiguous structure determination. Demonstrated on 5-aminosalicylic acid and S-naproxen (crystallite sizes of 50–55 nm), the approach provides a robust strategy for structure determination and represents a significant advance in the field of local-structure elucidation of poorly crystalline and nanocrystalline organic solids, where conventional diffraction techniques fail.
- 6 October 2026
- Published by Springer Nature in Communications Chemistry
- Vol. 9 (1) , 314
Open AccessPair distribution function “feat.” solid-state NMR in a combined approach for structure determination of organic molecules
Investigating the local structure of molecular solids is critical for establishing structure–property relationships and controlling material performance. Pair distribution function (PDF) analysis probes poorly crystalline, nanocrystalline and amorphous materials, yet its application to organic compounds remains limited. The recently developed Global-PDF-Fit method enables ab initio crystal structure determination of organic molecules, although intrinsic limitations can complicate unique structure identification. Here, we present a synergistic workflow integrating Global-PDF-Fit with solid-state NMR (SSNMR), demonstrating how restraints and intermolecular proximity constraints from 1D magic angle spinning (MAS) and 2D 1 H- 13 C double cross polarization and 1 H double quantum/ 1 H single quantum MAS SSNMR experiments reduce computational time, guide structure ranking, and increase reliability, enabling unambiguous structure determination. Demonstrated on 5-aminosalicylic acid and S-naproxen (crystallite sizes of 50–55 nm), the approach provides a robust strategy for structure determination and represents a significant advance in the field of local-structure elucidation of poorly crystalline and nanocrystalline organic solids, where conventional diffraction techniques fail.
- Open Access
- 30 June 2026
- Published by American Chemical Society (ACS) in Journal of the American Chemical Society
- Vol. 148 (27) , 28754-28763
GraPhAI: Neural Networks for Solving Centrosymmetric Crystal Structures
Crystal structure determination from low-resolution diffraction data is challenging due to the lack of a general-purpose method for solving the crystallographic phase problem. We have developed a graph neural network, GraPhAI, and propose a new and efficient diffraction data representation in a graph form for deep learning. The trained GraPhAI models are intended for ab initio phasing of down to 2 Å resolution data of typical unit-cell volume centrosymmetric crystal structures. The success rate is above 80% for structures containing an atom with Z ≥ 19 (e.g., metal-organic frameworks, coordination compounds, or inorganic structures). For purely organic crystal structures, the success so far is limited.
- 30 June 2026
- Published by American Chemical Society (ACS) in Journal of the American Chemical Society
- Vol. 148 (27) , 28754-28763
Open AccessGraPhAI: Neural Networks for Solving Centrosymmetric Crystal Structures
Crystal structure determination from low-resolution diffraction data is challenging due to the lack of a general-purpose method for solving the crystallographic phase problem. We have developed a graph neural network, GraPhAI, and propose a new and efficient diffraction data representation in a graph form for deep learning. The trained GraPhAI models are intended for ab initio phasing of down to 2 Å resolution data of typical unit-cell volume centrosymmetric crystal structures. The success rate is above 80% for structures containing an atom with Z ≥ 19 (e.g., metal-organic frameworks, coordination compounds, or inorganic structures). For purely organic crystal structures, the success so far is limited.
- Open Access
- 12 March 2026
- Published by American Chemical Society (ACS) in ChemRxiv
GraPhAI: Neural Networks for Solving Centrosymmetric Crystal Structures
Crystal structure determination from low resolution diffraction data is challenging due to lack of a general-purpose method for solving the crystallographic phase problem. We have developed graph neural network GraPhAI and propose a new and efficient diffraction data representation in a graph form for deep learning. The trained GraPhAI models are intended for ab initio phasing of down to 2 Å resolution data of typical unit cell volume centrosymmetric crystal structures. The success rate is above 80% for structures containing an atom with Z ≥ 19 (e.g. metal-organic frameworks, coordination compounds, or inorganic structures). For purely organic crystal structures, the success so far is limited.
- 12 March 2026
- Published by American Chemical Society (ACS) in ChemRxiv
Open AccessGraPhAI: Neural Networks for Solving Centrosymmetric Crystal Structures
Crystal structure determination from low resolution diffraction data is challenging due to lack of a general-purpose method for solving the crystallographic phase problem. We have developed graph neural network GraPhAI and propose a new and efficient diffraction data representation in a graph form for deep learning. The trained GraPhAI models are intended for ab initio phasing of down to 2 Å resolution data of typical unit cell volume centrosymmetric crystal structures. The success rate is above 80% for structures containing an atom with Z ≥ 19 (e.g. metal-organic frameworks, coordination compounds, or inorganic structures). For purely organic crystal structures, the success so far is limited.
- Open Access
- 1 January 2026
- Published by International Union of Crystallography (IUCr) in Acta Crystallographica Section A Foundations and Advances
- Vol. 82 (1) , 32-40
On artificial crystal structure generation for solving the phase problem with deep learning
We discuss and present approaches for generating artificial crystal structures for training neural networks to solve the phase problem. Structure generation is considered as a two-step process involving sampling unit-cell parameters and filling the unit cell with atoms. The former step includes generating lattice basis vectors from randomly sampled unit-cell volume. Apart from randomly placing atoms, we use database data to guide fast and scalable generation of molecule-like fragments. The recently developed neural network PhAI is then used as a benchmark and retrained with various sets of training data to assess how the corresponding models perform on experimental crystal structure data. We found a significant improvement in PhAI retrained on a new kind of artificial data to generalize the phase problem solution for larger unit-cell structures.
- 1 January 2026
- Published by International Union of Crystallography (IUCr) in Acta Crystallographica Section A Foundations and Advances
- Vol. 82 (1) , 32-40
Open AccessOn artificial crystal structure generation for solving the phase problem with deep learning
We discuss and present approaches for generating artificial crystal structures for training neural networks to solve the phase problem. Structure generation is considered as a two-step process involving sampling unit-cell parameters and filling the unit cell with atoms. The former step includes generating lattice basis vectors from randomly sampled unit-cell volume. Apart from randomly placing atoms, we use database data to guide fast and scalable generation of molecule-like fragments. The recently developed neural network PhAI is then used as a benchmark and retrained with various sets of training data to assess how the corresponding models perform on experimental crystal structure data. We found a significant improvement in PhAI retrained on a new kind of artificial data to generalize the phase problem solution for larger unit-cell structures.
- Open Access
- 4 February 2025
- Published by American Chemical Society (ACS) in Crystal Growth & Design
- Vol. 25 (4) , 1148-1154
Incommensurate Displacive and Occupational Modulation in Hexakis(Urea-O)Iron(III) Nitrate
- 4 February 2025
- Published by American Chemical Society (ACS) in Crystal Growth & Design
- Vol. 25 (4) , 1148-1154
Open AccessIncommensurate Displacive and Occupational Modulation in Hexakis(Urea-O)Iron(III) Nitrate
- Open Access
- 3 February 2025
- Published by Elsevier in Journal of Solid State Chemistry
- Vol. 345
Deciphering the commensurately modulated monoclinic phase of RbZnCl at low temperatures- 3 February 2025
- Published by Elsevier in Journal of Solid State Chemistry
- Vol. 345
Open AccessDeciphering the commensurately modulated monoclinic phase of RbZnCl at low temperatures - 24 January 2025
- Published by American Chemical Society (ACS) in Crystal Growth & Design
- Vol. 25 (3) , 593-602
Phase Transition in the Jumping Crystal l-Pyroglutamic Acid: Insights from Dynamic Quantum Crystallography and Spectroscopy
- 24 January 2025
- Published by American Chemical Society (ACS) in Crystal Growth & Design
- Vol. 25 (3) , 593-602
Phase Transition in the Jumping Crystal l-Pyroglutamic Acid: Insights from Dynamic Quantum Crystallography and Spectroscopy
- 24 August 2024
- Published by American Chemical Society (ACS) in Crystal Growth & Design
- Vol. 24 (18) , 7677-7685
Fifteen Solid Solutions of Four Thioxanthone Halogen Derivatives: Structures, Miscibility Limits, and Luminescence
- 24 August 2024
- Published by American Chemical Society (ACS) in Crystal Growth & Design
- Vol. 24 (18) , 7677-7685
Fifteen Solid Solutions of Four Thioxanthone Halogen Derivatives: Structures, Miscibility Limits, and Luminescence
- 2 August 2024
- Published by American Association for the Advancement of Science (AAAS) in Science
- Vol. 385 (6708) , 522-528
PhAI: A deep-learning approach to solve the crystallographic phase problem
X-ray crystallography provides a distinctive view on the three-dimensional structure of crystals. To reconstruct the electron density map, the complex structure factors F = F exp i ϕ of a sufficiently large number of diffracted reflections must be known. In a conventional experiment, only the amplitudes F are obtained, and the phases ϕ are lost. This is the crystallographic phase problem. In this work, we show that a neural network, trained on millions of artificial structure data, can solve the phase problem at a resolution of only 2 angstroms, using only 10 to 20% of the data needed for direct methods. The network works in common space groups and for modest unit-cell dimensions and suggests that neural networks could be used to solve the phase problem in the general case for weakly scattering crystals.
- 2 August 2024
- Published by American Association for the Advancement of Science (AAAS) in Science
- Vol. 385 (6708) , 522-528
PhAI: A deep-learning approach to solve the crystallographic phase problem
X-ray crystallography provides a distinctive view on the three-dimensional structure of crystals. To reconstruct the electron density map, the complex structure factors F = F exp i ϕ of a sufficiently large number of diffracted reflections must be known. In a conventional experiment, only the amplitudes F are obtained, and the phases ϕ are lost. This is the crystallographic phase problem. In this work, we show that a neural network, trained on millions of artificial structure data, can solve the phase problem at a resolution of only 2 angstroms, using only 10 to 20% of the data needed for direct methods. The network works in common space groups and for modest unit-cell dimensions and suggests that neural networks could be used to solve the phase problem in the general case for weakly scattering crystals.