1
|
Wei J, Liang J, Song J, Zhou P. YOLO-PBESW: A Lightweight Deep Learning Model for the Efficient Identification of Indomethacin Crystal Morphologies in Microfluidic Droplets. MICROMACHINES 2024; 15:1136. [PMID: 39337796 PMCID: PMC11433745 DOI: 10.3390/mi15091136] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Grants] [Track Full Text] [Subscribe] [Scholar Register] [Received: 08/10/2024] [Revised: 09/04/2024] [Accepted: 09/05/2024] [Indexed: 09/30/2024]
Abstract
Crystallization is important to the pharmaceutical, the chemical, and the materials fields, where the morphology of crystals is one of the key factors affecting the quality of crystallization. High-throughput screening based on microfluidic droplets is a potent technique to accelerate the discovery and development of new crystal morphologies with active pharmaceutical ingredients. However, massive crystal morphologies' datum needs to be identified completely and accurately, which is time-consuming and labor-intensive. Therefore, effective morphologies' detection and small-target tracking are essential for high-efficiency experiments. In this paper, a new improved algorithm YOLOv8 (YOLO-PBESW) for detecting indomethacin crystals with different morphologies is proposed. We enhanced its capability in detecting small targets through the integration of a high-resolution feature layer P2, and the adoption of a BiFPN structure. Additionally, in this paper, adding the EMA mechanism before the P2 detection head was implemented to improve network attention towards global features. Furthermore, we utilized SimSPPF to replace SPPF to mitigate computational costs and reduce inference time. Lastly, the CIoU loss function was substituted with WIoUv3 to improve detection performance. The experimental findings indicate that the enhanced YOLOv8 model attained advancements, achieving AP metrics of 93.3%, 77.6%, 80.2%, and 99.5% for crystal wire, crystal rod, crystal sheet, and jelly-like phases, respectively. The model also achieved a precision of 85.2%, a recall of 83.8%, and an F1 score of 84.5%, with a mAP of 87.6%. In terms of computational efficiency, the model's dimensions and operational efficiency are reported as 5.46 MB, and it took 12.89 ms to process each image with a speed of 77.52 FPS. Compared with state-of-the-art lightweight small object detection models such as the FFCA-YOLO series, our proposed YOLO-PBESW model achieved improvements in detecting indomethacin crystal morphologies, particularly for crystal sheets and crystal rods. The model demonstrated AP values that exceeded L-FFCA-YOLO by 7.4% for crystal sheets and 3.9% for crystal rods, while also delivering a superior F1-score. Furthermore, YOLO-PBESW maintained a lower computational complexity, with parameters of only 11.8 GFLOPs and 2.65 M, and achieved a higher FPS. These outcomes collectively demonstrate that our method achieved a balance between precision and computational speed.
Collapse
Affiliation(s)
- Jiehan Wei
- School of Mechatronic Engineering, Guangdong Polytechnic Normal University, Guangzhou 510665, China
| | - Jianye Liang
- School of Mechatronic Engineering, Guangdong Polytechnic Normal University, Guangzhou 510665, China
| | - Jun Song
- School of Mechatronic Engineering, Guangdong Polytechnic Normal University, Guangzhou 510665, China
| | - Peipei Zhou
- School of Mechatronic Engineering, Guangdong Polytechnic Normal University, Guangzhou 510665, China
| |
Collapse
|
2
|
Liu Z, Gu KK, Shelby ML, Gilbile D, Lyubimov AY, Russi S, Cohen AE, Narayanasamy SR, Botha S, Kupitz C, Sierra RG, Poitevin F, Gilardi A, Lisova S, Coleman MA, Frank M, Kuhl TL. A user-friendly plug-and-play cyclic olefin copolymer-based microfluidic chip for room-temperature, fixed-target serial crystallography. Acta Crystallogr D Struct Biol 2023; 79:944-952. [PMID: 37747292 PMCID: PMC10565732 DOI: 10.1107/s2059798323007027] [Citation(s) in RCA: 3] [Impact Index Per Article: 3.0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 02/09/2023] [Accepted: 08/07/2023] [Indexed: 09/26/2023] Open
Abstract
Over the past two decades, serial X-ray crystallography has enabled the structure determination of a wide range of proteins. With the advent of X-ray free-electron lasers (XFELs), ever-smaller crystals have yielded high-resolution diffraction and structure determination. A crucial need to continue advancement is the efficient delivery of fragile and micrometre-sized crystals to the X-ray beam intersection. This paper presents an improved design of an all-polymer microfluidic `chip' for room-temperature fixed-target serial crystallography that can be tailored to broadly meet the needs of users at either synchrotron or XFEL light sources. The chips are designed to be customized around different types of crystals and offer users a friendly, quick, convenient, ultra-low-cost and robust sample-delivery platform. Compared with the previous iteration of the chip [Gilbile et al. (2021), Lab Chip, 21, 4831-4845], the new design eliminates cleanroom fabrication. It has a larger imaging area to volume, while maintaining crystal hydration stability for both in situ crystallization or direct crystal slurry loading. Crystals of two model proteins, lysozyme and thaumatin, were used to validate the effectiveness of the design at both synchrotron (lysozyme and thaumatin) and XFEL (lysozyme only) facilities, yielding complete data sets with resolutions of 1.42, 1.48 and 1.70 Å, respectively. Overall, the improved chip design, ease of fabrication and high modifiability create a powerful, all-around sample-delivery tool that structural biologists can quickly adopt, especially in cases of limited sample volume and small, fragile crystals.
Collapse
Affiliation(s)
- Zhongrui Liu
- Department of Chemical Engineering, University of California at Davis, Davis, CA 95616, USA
| | - Kevin K. Gu
- Department of Chemical Engineering, University of California at Davis, Davis, CA 95616, USA
| | - Megan L. Shelby
- Biosciences and Biotechnology Division, Lawrence Livermore National Laboratory, Livermore, CA 94550, USA
| | - Deepshika Gilbile
- Department of Chemical Engineering, University of California at Davis, Davis, CA 95616, USA
| | - Artem Y. Lyubimov
- Stanford Synchrotron Radiation Lightsource, SLAC National Accelerator Laboratory, Menlo Park, CA 94025, USA
| | - Silvia Russi
- Stanford Synchrotron Radiation Lightsource, SLAC National Accelerator Laboratory, Menlo Park, CA 94025, USA
| | - Aina E. Cohen
- Stanford Synchrotron Radiation Lightsource, SLAC National Accelerator Laboratory, Menlo Park, CA 94025, USA
| | - Sankar Raju Narayanasamy
- Biosciences and Biotechnology Division, Lawrence Livermore National Laboratory, Livermore, CA 94550, USA
| | - Sabine Botha
- Department of Physics, Arizona State University, Tempe, AZ 85287, USA
| | - Christopher Kupitz
- Linac Coherent Light Source, SLAC National Accelerator Laboratory, Menlo Park, CA 94025, USA
| | - Raymond G. Sierra
- Linac Coherent Light Source, SLAC National Accelerator Laboratory, Menlo Park, CA 94025, USA
| | - Fredric Poitevin
- Linac Coherent Light Source, SLAC National Accelerator Laboratory, Menlo Park, CA 94025, USA
| | - Antonio Gilardi
- Linac Coherent Light Source, SLAC National Accelerator Laboratory, Menlo Park, CA 94025, USA
| | - Stella Lisova
- Linac Coherent Light Source, SLAC National Accelerator Laboratory, Menlo Park, CA 94025, USA
| | - Matthew A. Coleman
- Biosciences and Biotechnology Division, Lawrence Livermore National Laboratory, Livermore, CA 94550, USA
- Department of Radiation Oncology, School of Medicine, University of California at Davis, Sacramento, CA 95817, USA
| | - Matthias Frank
- Biosciences and Biotechnology Division, Lawrence Livermore National Laboratory, Livermore, CA 94550, USA
- Department of Biochemistry and Molecular Medicine, School of Medicine, University of California at Davis, Sacramento, CA 95817, USA
| | - Tonya L. Kuhl
- Department of Chemical Engineering, University of California at Davis, Davis, CA 95616, USA
| |
Collapse
|
3
|
Thompson MC. Combining temperature perturbations with X-ray crystallography to study dynamic macromolecules: A thorough discussion of experimental methods. Methods Enzymol 2023; 688:255-305. [PMID: 37748829 DOI: 10.1016/bs.mie.2023.07.008] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 09/27/2023]
Abstract
Temperature is an important state variable that governs the behavior of microscopic systems, yet crystallographers rarely exploit temperature changes to study the structure and dynamics of biological macromolecules. In fact, approximately 90% of crystal structures in the Protein Data Bank were determined under cryogenic conditions, because sample cryocooling makes crystals robust to X-ray radiation damage and facilitates data collection. On the other hand, cryocooling can introduce artifacts into macromolecular structures, and can suppress conformational dynamics that are critical for function. Fortunately, recent advances in X-ray detector technology, X-ray sources, and computational data processing algorithms make non-cryogenic X-ray crystallography easier and more broadly applicable than ever before. Without the reliance on cryocooling, high-resolution crystallography can be combined with various temperature perturbations to gain deep insight into the conformational landscapes of macromolecules. This Chapter reviews the historical reasons for the prevalence of cryocooling in macromolecular crystallography, and discusses its potential drawbacks. Next, the Chapter summarizes technological developments and methodologies that facilitate non-cryogenic crystallography experiments. Finally, the chapter discusses the theoretical underpinnings and practical aspects of multi-temperature and temperature-jump crystallography experiments, which are powerful tools for understanding the relationship between the structure, dynamics, and function of proteins and other biological macromolecules.
Collapse
Affiliation(s)
- Michael C Thompson
- Department of Chemistry and Biochemistry, University of California, Merced, Merced, CA, United States.
| |
Collapse
|
4
|
Saha S, Özden C, Samkutty A, Russi S, Cohen A, Stratton MM, Perry SL. Polymer-based microfluidic device for on-chip counter-diffusive crystallization and in situ X-ray crystallography at room temperature. LAB ON A CHIP 2023; 23:2075-2090. [PMID: 36942575 PMCID: PMC10631519 DOI: 10.1039/d2lc01194h] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [MESH Headings] [Grants] [Track Full Text] [Subscribe] [Scholar Register] [Indexed: 05/05/2023]
Abstract
Proteins are long chains of amino acid residues that perform a myriad of functions in living organisms, including enzymatic reactions, signalling, and maintaining structural integrity. Protein function is determined directly by the protein structure. X-ray crystallography is the primary technique for determining the 3D structure of proteins, and facilitates understanding the effects of protein structure on function. The first step towards structure determination is crystallizing the protein of interest. We have developed a centrifugally-actuated microfluidic device that incorporates the fluid handling and metering necessary for protein crystallization. Liquid handling takes advantage of surface forces to control fluid flow and enable metering, without the need for any fluidic or pump connections. Our approach requires only the simple steps of pipetting the crystallization reagents into the device followed by either spinning or shaking to set up counter-diffusive protein crystallization trials. The use of thin, UV-curable polymers with a high level of X-ray transparency allows for in situ X-ray crystallography, eliminating the manual handling of fragile protein crystals and streamlining the process of protein structure analysis. We demonstrate the utility of our device using hen egg white lysozyme as a model system, followed by the crystallization and in situ, room temperature structural analysis of the hub domain of calcium-calmodulin dependent kinase II (CaMKIIβ).
Collapse
Affiliation(s)
- Sarthak Saha
- Department of Chemical Engineering, University of Massachusetts Amherst, MA 01003, USA.
| | - Can Özden
- Department of Biochemistry and Molecular Biology, University of Massachusetts Amherst, MA 01003, USA
| | - Alfred Samkutty
- Department of Biochemistry and Molecular Biology, University of Massachusetts Amherst, MA 01003, USA
| | - Silvia Russi
- SLAC National Accelerator Laboratory, Menlo Park, CA 94025, USA
| | - Aina Cohen
- SLAC National Accelerator Laboratory, Menlo Park, CA 94025, USA
| | - Margaret M Stratton
- Department of Biochemistry and Molecular Biology, University of Massachusetts Amherst, MA 01003, USA
| | - Sarah L Perry
- Department of Chemical Engineering, University of Massachusetts Amherst, MA 01003, USA.
| |
Collapse
|
5
|
Skaist Mehlman T, Biel JT, Azeem SM, Nelson ER, Hossain S, Dunnett L, Paterson NG, Douangamath A, Talon R, Axford D, Orins H, von Delft F, Keedy DA. Room-temperature crystallography reveals altered binding of small-molecule fragments to PTP1B. eLife 2023; 12:84632. [PMID: 36881464 PMCID: PMC9991056 DOI: 10.7554/elife.84632] [Citation(s) in RCA: 20] [Impact Index Per Article: 20.0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 11/15/2022] [Accepted: 02/12/2023] [Indexed: 03/08/2023] Open
Abstract
Much of our current understanding of how small-molecule ligands interact with proteins stems from X-ray crystal structures determined at cryogenic (cryo) temperature. For proteins alone, room-temperature (RT) crystallography can reveal previously hidden, biologically relevant alternate conformations. However, less is understood about how RT crystallography may impact the conformational landscapes of protein-ligand complexes. Previously, we showed that small-molecule fragments cluster in putative allosteric sites using a cryo crystallographic screen of the therapeutic target PTP1B (Keedy et al., 2018). Here, we have performed two RT crystallographic screens of PTP1B using many of the same fragments, representing the largest RT crystallographic screens of a diverse library of ligands to date, and enabling a direct interrogation of the effect of data collection temperature on protein-ligand interactions. We show that at RT, fewer ligands bind, and often more weakly - but with a variety of temperature-dependent differences, including unique binding poses, changes in solvation, new binding sites, and distinct protein allosteric conformational responses. Overall, this work suggests that the vast body of existing cryo-temperature protein-ligand structures may provide an incomplete picture, and highlights the potential of RT crystallography to help complete this picture by revealing distinct conformational modes of protein-ligand systems. Our results may inspire future use of RT crystallography to interrogate the roles of protein-ligand conformational ensembles in biological function.
Collapse
Affiliation(s)
- Tamar Skaist Mehlman
- Structural Biology Initiative, CUNY Advanced Science Research CenterNew YorkUnited States
- PhD Program in Biochemistry, CUNY Graduate CenterNew YorkUnited States
| | - Justin T Biel
- Department of Bioengineering and Therapeutic Sciences, University of California, San FranciscoSan FranciscoUnited States
| | - Syeda Maryam Azeem
- Structural Biology Initiative, CUNY Advanced Science Research CenterNew YorkUnited States
| | | | - Sakib Hossain
- Structural Biology Initiative, CUNY Advanced Science Research CenterNew YorkUnited States
| | - Louise Dunnett
- Diamond Light SourceDidcotUnited Kingdom
- Research Complex at Harwell, Harwell Science and Innovation CampusDidcotUnited Kingdom
| | | | - Alice Douangamath
- Diamond Light SourceDidcotUnited Kingdom
- Research Complex at Harwell, Harwell Science and Innovation CampusDidcotUnited Kingdom
| | | | | | - Helen Orins
- Structural Biology Initiative, CUNY Advanced Science Research CenterNew YorkUnited States
| | - Frank von Delft
- Diamond Light SourceDidcotUnited Kingdom
- Research Complex at Harwell, Harwell Science and Innovation CampusDidcotUnited Kingdom
- Centre for Medicines Discovery, Nuffield Department of Medicine, University of OxfordOxfordUnited Kingdom
- Department of Biochemistry, University of JohannesburgJohannesburgSouth Africa
| | - Daniel A Keedy
- Structural Biology Initiative, CUNY Advanced Science Research CenterNew YorkUnited States
- Department of Chemistry and Biochemistry, City College of New YorkNew YorkUnited States
- PhD Programs in Biochemistry, Biology, and Chemistry, CUNY Graduate CenterNew YorkUnited States
| |
Collapse
|