Home
PDBCharges is a web application providing partial atomic charges of protein structures from the Protein Data Bank. The charges are computed by the semiempirical quantum mechanical methods GFN1-xTB and reproduce the PBE0/TZVP/CM5 charges. Before computation of the charges, hydrogens are added to the structure by Hydride and MoleculeKit at pH 7.2. The positions of the added hydrogens are also optimized using the GFN-FF force field.
The main page allows uploading the chosen structure from PDB. The structure must be defined by its PDB ID:
The PDB ID of the protein can be obtained directly from PDB. Here, you can search for your protein. Examples of PDB IDs, which are accepted by PDBCharges: 1alf, 2pws, or 6wlv. PDB ID can be lowercase or uppercase.
By clicking on the Get charges button, the user is redirected to the results page for that protein.
The Main page also offers three use cases – proteins, which partial atomic charges provide interesting information for the research community.
The result page includes visualisation of the charged structure and a download data section.
This information consists of the PDB ID of the structure, its number of atoms, total charge and number of atoms for which the partial atomic charge calculation failed (see Limitations). An example of the information is the following:
PDBCharges integrates Mol* viewer to show the calculated charges. The user can select three different visualisation modes:
- Cartoon, in which the colour of individual amino acids is determined by the sum of charges on its atoms.
- Surface, in which partial atomic charges of the nearest atom colour the protein surface.
- Ball & Stick, in which the individual atoms are coloured according to their partial atomic charges.
It is also possible to hide water molecules (Show water) for a clearer view of protein structure. This is especially recommended for surface visualization of protein molecules.
Figure 1: Visualisation of the molecule with PDB entry 1uzv in the surface mode with (left) and without (right) water molecule.
Users can select from various methods to colour the molecule to suit their specific application needs.
- Structure (chain ID), in which colour is set by chain ID.
- Structure (uniform), in which colour is set by element symbol.
- Charges (relative), in which colour is set according to the highest absolute value of a computed charge.
- Charges (absolute), in which colour is set by the user, who specifies the range.
The Highlight charge differences option enables users to visually identify and compare areas in the molecule with similar charges. Recommended for cartoon models.
The redder the colour, the more negative the charge, and the bluer the colour, the more positive the charge. Atoms that did not converge using the QM partial atomic charge calculation are coloured green.
A panel displaying a sequence is at the top of the Mol* viewer. The panel always displays only the sequence for the selected chain, which can be set using the chain ID. If an atom in a residue or ligand has a warning that the residue or ligand will turn red.
Figure 2: Visualisation of the molecule with PDB entry 1uzv. The sequence is displayed at the top of the Mol* viewer.
The Show Residual Warnings button allows users to display a table with warnings related to optimization and atomic charge calculations. This table consists of five rows. The columns labelled "Chain ID," "Residue ID," and "Residue Name" indicate the specific location of each issue. There is also a "Warnings" row that provides more detailed information about each problem. Additionally, users can click the Show button next to each warning to focus on the problematic area within the molecule. It is also possible to use the search function, which is located above the table, to allow a quick search in the results.
The Download charges and protonated structure button allows the user to download a ZIP file containing two files, specifically:
- mmCIF file contains coordinates of the protonated structure and partial atomic charges. Details regarding partial charges are provided at the end of the mmCIF file. This section contains information about the calculation methods, atom identification, and the partial atomic charges of the atoms within the molecule.
loop_
_sb_ncbr_partial_atomic_charges_meta.id # id of the charges
_sb_ncbr_partial_atomic_charges_meta.type # type of the charges
_sb_ncbr_partial_atomic_charges_meta.method # calculation method name
1 'QM' 'GFN1-xTB/CM5 (with cutoff)'
loop_
_sb_ncbr_partial_atomic_charges.type_id # id of the charges (pointer to _sb_ncbr_partial_atomic_charges_meta.id)
_sb_ncbr_partial_atomic_charges.atom_id # atom id (pointer to _atom_site.id)
_sb_ncbr_partial_atomic_charges.charge # partial atomic charge
1 1 -0.6462
1 2 0.0888
1 3 0.4992
...
- Json file containing possible warning messages for individual residues. These warnings alert the user that the structure or calculated partial atomic charges may be incorrect. Details regarding warnings for partial charges calculation are provided in the JSON file. This section includes details about residue identification and warnings encountered during the process.
[
{
"chain_id": "A",
"residue_id": 1,
"residue_name": "ABU",
"warning": "Atom(s) O were added by PDBFixer. Charge calculation failed for atom(s) O."
},
{
"chain_id": "A",
"residue_id": 7,
"residue_name": "ACA",
"warning": "Atom(s) O were added by PDBFixer. Mapping of formal charges from Dimorphite-DL and CCD to residue failed and therefore the residue is left neutral."
}
]
-
TXT file contains charges of atoms present in the protonated structure. Individual values are separated by spaces. The charges are ordered by indices of the atoms for which they were computed.
-
PQR file contains information about the coordinates and charges of the protonated structure. It is similar to PDB, with the two last columns changed to partial atomic charge and atomic radius.
ATOM 1 N MET 1 -23.494 -19.088 -2.588 -0.7600 2.0000
ATOM 2 CA MET 1 -22.509 -19.864 -3.374 -0.1289 2.0000
ATOM 3 C MET 1 -21.922 -18.915 -4.401 0.6910 1.7000
ATOM 4 CB MET 1 -23.176 -21.067 -4.061 -0.4312 2.0000
ATOM 5 O MET 1 -22.699 -18.157 -4.963 -0.6850 1.4000
RDKit
The RDKit is a collection of cheminformatics and machine-learning software. One of the advantages of using RDKit for the PDBcharges tool is its extensive range of functions specifically designed for working with small molecules, such as ligands, as well as chemical bonds. However, its functionality for handling PDB files is limited, and it does not support mmCIF files. Documentation is available on the RDKit website and on GitHub.
Biopython
Biopython is a set of freely available tools for bioinformatical computation (Cock2009). The main benefit of using Biopython for the PDBcharges tool is its extensive range of functions for working with biomolecules. On the other hand, it does not support the analysis of chemical bonds. Documentation is available on the Biopython website, and the source code is available on GitHub.
GEMMI
Gemmi is an open-source library that includes a set of programs primarily designed for structural biology, specifically in macromolecular crystallography. Gemmi is a joint project of Global Phasing Ltd and CCP4 (Wojdyr2022). The advantage of using it for the tool is its advanced functionality for writing and editing mmCIF files. Documentation is available on the GEMMI website, and the source code is available on GitHub.
Hydride
Hydride is an easy-to-use program and library built on top of the Biotite library. It adds missing hydrogen atoms to molecular models based on known bond lengths and angles. It does not require force-field parameters for the specific molecules, it can be used for adding hydrogen atoms to almost any organic molecule - from small ligands to large protein complexes (Kunzmann2022). This flexibility is one reason it's used by PDBcharges. However, for accurate assignment of hydrogens, formal charges must be provided, as Hydride cannot calculate them correctly. Documentation is available on the Hydride website, and the source code is available on GitHub.
PDBFixer
PDBFixer is an easy-to-use application for fixing problems in Protein Data Bank files in preparation for simulating them. PDBFixer is built on top of the OpenMM library (Eastman2017). The main benefit is that PDBFixer primarily focuses on eliminating common errors in protein structures that could lower the accuracy of the calculation of partial atomic charges. More information about the application is available on GitHub and in the manual.
Dimorphite-DL
Dimorphite-DL is a fast, accurate, accessible, and modular open-source program designed to enumerate ionisation states of small molecules. It adds hydrogen atoms to molecular representations as appropriate for a user-specified pH range (Ropp2019). The primary advantage of this tool is its ability to assign formal charges to molecules universally. However, it is limited to only accepting input in SMILES format. More information about the application is available on GitHub and the Dimorphite-DL website.
MoleculeKit
Moleculekit is a library that offers object-oriented classes and methods for manipulating biomolecular structures (Doerr2016). It is built on top of pdb2pqr/PROPKA and has two significant advantages that led to its selection for the PDBcharges tool: it correctly adds hydrogens to proteins and accounts ligands when calculating protonation states. The downside is that it can't add hydrogens to ligands, so ligands must contain hydrogens before MoleculeKit calculation. Documentation is available on the Moleculekit website, and the source code for Moleculekit is available on GitHub.
XTB
XTB is a semiempirical quantum chemistry software package based on the extended tight-binding (xTB) methods. It is designed to provide efficient and accurate approximate quantum mechanical calculations for molecular systems (Bannwarth2021). One of the main advantages of using it for PDBcharges is its high configurability, offering several levels of theory, including GFN-FF and GFN1-xTB, which were used in the workflow for PDBcharges. Documentation is available on the XTB website, and the source code for xtb is available on GitHub.
Open Babel
Open Babel is an open, collaborative project allowing search, convert, analyze, or store data from molecular modelling, chemistry, solid-state materials, biochemistry, or related areas (O'Boyle2011). Its support of a large number of chemo/bioinformatics formats is a great advantage for the PDBcharges tool. More information about the tool is available on the Open Babel website, and the source code for xtb is available on GitHub.
GFN-FF
GFN-FF is a completely automated partially polarizable generic physics-based force-field for the accurate description of structures and dynamics of large molecules across the periodic table. It combines the speed of classical force fields with quantum mechanical accuracy (Spicher2020). We have several advantages why we chose it for PDBcharges and that is because it is accurate, physically based, fast, versatile and derived from QM.
GFN1-xTB
GFN1-xTB is a semiempirical quantum mechanical method classified as an extended tight-binding model. It is designed to efficiently and accurately predict molecular and periodic system properties, including geometries, vibrational frequencies, and non-covalent interactions, covering most periodic table elements up to radon (Z = 86) (Grimme2017). We chose GFN1-xTB for PDBcharges because it is a fast and versatile semiempirical quantum mechanics method.
Cover
Cover is a divide-and-conquer complexity reduction algorithm for calculating partial atomic charges (Raček2020). The time required for these calculations increases polynomially with the number of atoms in a molecule. Therefore, computing the partial atomic charges for large structures, such as proteins, can be unfeasible due to computational complexity. The cover approach reduces computational time by dividing the molecule into smaller, overlapping substructures of approximately equal size. Partial atomic charges are then calculated separately for each of these smaller parts. The cover approach replaces an extensive computation with multiple shorter ones with approximately constant computation time. It means that the polynomial relationship between computation time and the number of atoms is transformed into a linear relationship. For these benefits, we employed the Cover approach in our PDBCharges workflow.
Chemical Component Dictionary
The Chemical Component Dictionary (CCD) is an external reference file describing all residue and small molecule components found in PDB entries. This dictionary includes comprehensive chemical information for standard and modified amino acids and nucleotides, as well as small molecule ligands and solvent molecules (Berman2003). The significant advantage is that it includes a pattern for all ligands in PDB, but it is important to note that formal charges are not always assigned correctly. More information about the dictionary is available on the PDB/CCD website.
The calculation of partial atomic charges has five phases:
Correction of the structure
In the first step, the PDBFixer and OpenMM libraries identify and correct structural issues that could degrade the accuracy of partial atomic charge calculations. Specifically, this process involves adding missing heavy atoms to the structure and selecting one position for atoms with multiple alternate positions listed. The PDBFixer library detects and adds missing atoms by comparing the residues in the repaired structure with the library templates. A template describes a residue: what atoms it contains, how they are bonded to each other, and what conformation they take on. After detection, the missing atoms are added to the structure according to the template and then their position is optimized by minimizing their energy. If the protein structure contains hydrogens, they are removed in this step to ensure consistency in the results.
Adding hydrogens to hetero-residues
In the second step, hydrogen is added to the hetero-residue using the Hydride library. To accurately assign hydrogen to hetero-residues, it is essential to determine the protonation state of their functional groups. However, the Hydride library does not determine protonation states on its own. The protonation states are, therefore, obtained from the CCD. However, some functional groups within the CCD lack assigned protonation states. The Dimorphite-DL library provides formal charges based on the CCD to address this. The dimorphite-DL library can only work with structures in SMILES format. Therefore, the RDKit library maps the SMILES format charges to the PDB file. Once we have the information on the protonation states, the Hydride library assigns the hydrogens to the hetero-residues for physiological pH 7.2.
Adding hydrogens to standard residues
The MoleculeKit library, built on the pdb2pqr/PROPKA3 library, is designed to add hydrogen atoms to standard residues. MoleculeKit takes into account any non-protein, non-nucleic molecules during the calculation of formal charges and the addition of hydrogens. However, it is important to note that MoleculeKit cannot add hydrogens to hetero-residues. For this reason, hydrogens must be added to hetero-residues in a previous step of workflow. Hydrogens are added for physiological pH 7.2.
Optimisation of hydrogens
The added hydrogens are then optimised by the GFN-Force-Field from XTB software. The calculation of the partial atomic charges is sensitive to the quality of the structure, and thus, the optimisation of the hydrogen positions improves the accuracy of the resulting charges. Moreover, hydrogens are added to the structure by two different libraries (Hydride, MoleculeKit), and the optimisation unifies how hydrogen is placed. A cover approach is employed to speed up the optimisation of hydrogens. For each heavy atom, a substructure is defined that includes atoms within a range of 6 Å to 12 Å, ensuring that only the bonds between two carbon atoms are broken. Hydrogens are added to these broken C-C bonds. The optimisation of the hydrogen atoms that are at most 3 Å away from a given heavy atom is then conducted sequentially on these substructures. The creation of substructures is implemented using RDKit and Biotitle libraries.
Calculation of charges
The calculation of partial atomic charges is performed using the semi-empirical quantum mechanical method GFN1-xTB from XTB software. The GFN1-xTB method reproduces the PBE0/TZVP/CM5 charges (Grimme2016). Although GFN1-xTB is one of the fastest quantum chemical methods, calculating large molecules such as proteins is time-consuming. To overcome the issue of computation complexity, we use the Cover approach. In this approach, a substructure is defined for each heavy atom, considering atoms within a range of 6 Å to 12 Å, ensuring that only the bonds between two carbon atoms are broken. Hydrogens are added to these broken C-C bonds and then partial atomic charges are calculated for the resulting substructure. From that calculation, the partial atomic charges of the given heavy atom and all hydrogens bonded to that heavy atom are included in the final result. The creation of substructures is implemented using the RDKit and Biotite libraries.
The computing part has its own repository.
Phospholipase A2 (PLA2) is a water-soluble enzyme responsible for the hydrolysis of the sn-2 ester bond of the phospholipid substrate (Six2000). PLA2 can be sourced from plants, mammals, snakes, and bee venoms. Depending on their cellular location, these enzymes are classified into three categories: cytosolic (cPLA2), calcium-independent (iPLA2), and secretory (sPLA2) (Schaloske2006).
The products of hydrolysis, which include free fatty acids and lysophospholipids, can disrupt cellular membranes and trigger inflammatory responses. Free fatty acids like arachidonic acid and oleic acid are important not only for energy storage but also because they function as second messengers. Arachidonic acid, in particular, is a precursor to eicosanoids, potent mediators of inflammation and signal transduction (Six2000). Additionally, PLA2s are involved in membrane remodelling, a vital process for maintaining cellular homeostasis. This involves PLA2-catalyzed deacylation followed by reacylation through transacylases or acyltransferases, leading to the replacement of fatty acid moieties at the sn-2 position of membrane glycerophospholipids.
In studies, snake venom PLA2 has been shown to induce various pharmacological and toxic effects. These effects include presynaptic and postsynaptic neurotoxicity, myotoxicity, edema formation, hemorrhage, and anticoagulation. Some toxins exhibit a single effect, while others may induce multiple pharmacological responses (Gowda1994).
The carboxyl group of ibuprofen interacts, creating an electrostatic bond to lysin (LYS 60) in phospholipase A2 and inhibiting its function. Based on this knowledge, we can develop therapeutics that can reduce the inflammatory reactions associated with snakebites.
Pseudomonas aeruginosa is a gram-negative bacterium and an opportunistic pathogen that can infect almost every human tissue, especially when the immune system is compromised. It is commonly found in immunocompromised patients and is frequently associated with nosocomial infections. The bacteria colonise patients suffering from a number of chronic lung diseases, particularly those under assisted ventilation. Their infections are frequently fatal for cystic fibrosis patients (Mitchell2005, Mitchell2002).
Carbohydrates play a significant role in infections since they are a target for bacterial binding through virulence factors such as adhesins, present on pili or flagella, and soluble lectins. The tetrameric lectin first identified and characterised from the cytoplasm of P. aeruginosa has been shown to exist in large quantities on the outer membrane of the bacteria and is involved in biofilm formation (Adam2007).
Typically, interactions between proteins and carbohydrates involve a combination of hydrogen bonds and hydrophobic contacts. However, PA-IIL stands out because it primarily interacts with fucose through coordination with calcium ions and hydrogen bonds (Adam2007). PA-IIL exhibits a high affinity for L-fucose. In the PA-IIL/fucose complex, the fucose residue locks onto a pair of calcium ions with three hydroxyl groups (chain C: FUC 1118) participating in coordinating these cations. This interaction results in a stable complex through extensive charge delocalisation. Unlike most protein-carbohydrate interactions, PA-IIL relies on ionic and coordination bonds with minimal hydrophobic bonds, presenting the structural role of calcium ions in stabilising the binding site. Consequently stabilisinging molecules involved in the pathogen-specific recognition of modified host cell components is of therapeutic value (Mitchell2002, Adam2007). Understanding and rationalising the structural and thermodynamic basis of interaction between lectins from pathogens and carbohydrates opens the door to designing high-affinity inhibitors. Furthermore, a detailed analysis of these interactions and insights into crucial amino acids responsible for affinity and specificity can lead to targeted protein engineering of carbohydrate-binding proteins, allowing for fine-tuning of their properties for applications in biotechnology, pharmacology, and bioanalysis (Adam2007).
Potassium channels are found in nearly all living cells. They facilitate the movement of potassium ions across cell membranes and play a crucial role in various cellular functions, including neuronal firing, muscle contraction, volume regulation, and hormone secretion. The kidneys are particularly important for potassium (K+) secretion, maintaining K+ homeostasis (Reyes1998).
TWIK-related acid-sensitive K+ channel-2 (TASK-2) is a pH-sensitive member of the two-pore domain K+ (K2P) channel family. It is essential for several biological processes by regulating the balance of K+ across cell membranes (Zhang2022). TASK2 is widely expressed in various cell types, including neurons, immune cells, chondrocytes, and epithelial cells in organs such as the kidney. This channel has been implicated in chemosensation, volume regulation, and ionic homeostasis. An increase in TASK2 expression is associated with breast cancer proliferation, while its loss of function is linked to Balkan endemic nephropathy. The physiological roles of TASK2 have been related to its modulation by changes in intracellular and extracellular pH. In the neurons of the retrotrapezoid nucleus, inhibition of TASK2 by protons from the extracellular and/or intracellular environment leads to depolarization of the cell, which increases spike frequency and enhances respiration. In the kidney proximal tubule, activation of TASK2 by extracellular alkalization that results from electrogenic bicarbonate secretion hyperpolarizes the cell to support further bicarbonate efflux (Li2020).
The transmembrane regions of TASK2 are characterised by their non-polar nature and lack of charge, distinguishing them from the intracellular and extracellular domains of the protein.
- The MoleculeKit library is based on the PDB2PQR/PROPKA3 library, which can only handle structures in PDB format. If a structure is not available in this format, users will be notified on the landing page that charge calculations cannot be performed.
- The charges are calculated only for the first model if multiple models are available for a structure (typically structures determined by NMR).
- Several issues may occur during the calculation, such as non-convergence of hydrogen optimization or quantum-mechanical charge calculation. However, due to the Cover approach, these problems are only local, and the results for the rest of the structure remain unaffected. Users are informed of any problems on the results page. Information about the issues is also included in the downloaded partial atomic charge files in JSON format.
| OS | Version | Chrome | Firefox | Edge | Safari |
|---|---|---|---|---|---|
| Linux | Ubuntu 24.04 | 131 | 133 | 131 | n/a |
| Windows | Windows 11 | 131 | 133 | 133 | n/a |
| MacOS | Sonoma 14.2 | 130 | n/a | n/a | 17.2 |
If you encounter an error or have an idea for an improvement, please send a report to Ondrej Schindler (ondrej.schindler@mail.muni.cz) or open a GitHub issue. Thank you!
- (Cock2009) Cock, P.J.A. et al. Biopython: freely available Python tools for computational molecular biology and bioinformatics. Bioinformatics 2009 Jun 1; 25(11) 1422-3. 10.1093/bioinformatics/btp163 pmid:19304878
- (Wojdyr2022) Wojdyr, M.. GEMMI: A library for structural biology. Journal of Open Source Software, 7(73), 4200, 10.21105/joss.04200
- (Kunzmann2022) Kunzmann, P. Anter, J. M. & Hamacher, K. Algorithms Mol. Biol. 2022;17:7. 10.1186/s13015-022-00215-x
- (Ropp2019) Ropp PJ, Kaminsky JC, Yablonski S, Durrant JD. Dimorphite-DL: An open-source program for enumerating the ionization states of drug-like small molecules. J Cheminform 2019;11:14. doi:10.1186/s13321-019-0336-9
- (Doerr2016) Doerr S., Matthew J. Harvey, Frank Noé, and Gianni De Fabritiis. HTMD: High-throughput molecular dynamics for molecular discovery. Journal of Chemical Theory and Computation, 2016, 12 (4), pp 1845–1852. doi:10.1021/acs.jctc.6b00049
- (Eastman2017) Eastman P, Swails J, Chodera JD, McGibbon RT, Zhao Y, Beauchamp KA, Wang LP, Simmonett AC, Harrigan MP, Stern CD, Wiewiora RP, Brooks BR, Pande VS. OpenMM 7: Rapid development of high performance algorithms for molecular dynamics. PLoS Comput Biol. 2017 Jul 26;13(7):e1005659. doi: 10.1371/journal.pcbi.1005659. PMID: 28746339; PMCID: PMC5549999
- (Spicher2020) Spicher S, Grimme S. Robust Atomistic Modeling of Materials, Organometallic, and Biochemical Systems. Angew Chem Int Ed Engl. 2020 Sep 1;59(36):15665-15673. doi: 10.1002/anie.202004239. Epub 2020 May 18. PMID: 32343883; PMCID: PMC7267649
- (Grimme2017) Grimme S, Bannwarth C, Shushkov P. A Robust and Accurate Tight-Binding Quantum Chemical Method for Structures, Vibrational Frequencies, and Noncovalent Interactions of Large Molecular Systems Parametrized for All spd-Block Elements (Z = 1-86). J Chem Theory Comput. 2017;13(5):1989-2009. doi:10.1021/acs.jctc.7b00118
- (Bannwarth2021) Bannwarth C, Caldeweyher E, Ehlert S, et al. Extended tight-binding quantum chemistry methods. WIREs Comput Mol Sci. 2021; 11:e1493. doi.org/10.1002/wcms.1493
- (O'Boyle2011) O'Boyle NM, Banck M, James CA, Morley C, Vandermeersch T, Hutchison GR. Open Babel: An open chemical toolbox. J Cheminform. 2011 Oct 7;3:33. doi: 10.1186/1758-2946-3-33. PMID: 21982300; PMCID: PMC3198950
- (Berman2003) Berman H, Henrick K, Nakamura H. Announcing the worldwide Protein Data Bank. Nat Struct Biol. 2003;10(12):980. doi:10.1038/nsb1203-980
- (Raček2020) Raček T, Schindler O, Toušek D, Horský V, Berka K, Koča J, Svobodová R. Atomic Charge Calculator II: web-based tool for the calculation of partial atomic charges. Nucleic Acids Res. 2020 Jul 2;48(W1):W591-W596. doi: 10.1093/nar/gkaa367. PMID: 32402071; PMCID: PMC7319571
- (Grimme2016) Grimme S, Bannwarth C. Ultra-fast computation of electronic spectra for large systems by tight-binding based simplified Tamm-Dancoff approximation (sTDA-xTB). J Chem Phys. 2016;145(5):054103. doi:10.1063/1.4959605
- (Six2000) Six DA, Dennis EA. The expanding superfamily of phospholipase A(2) enzymes: classification and characterization. Biochim Biophys Acta. 2000;1488(1-2):1-19. doi:10.1016/s1388-1981(00)00105-0
- (Schaloske2006) Schaloske RH, Dennis EA. The phospholipase A2 superfamily and its group numbering system. Biochim Biophys Acta. 2006;1761(11):1246-1259. doi:10.1016/j.bbalip.2006.07.011
- (Gowda1994) Gowda TV, Middlebrook JL. Monoclonal antibodies to VRV-PL-VIIIa, a basic multitoxic phospholipase A2 from Vipera russelli venom. Toxicon. 1994;32(8):955-964. doi:10.1016/0041-0101(94)90374-3
- (Mitchell2005) Mitchell E, Sabin C, Snajdrová L, et al. High affinity fucose binding of Pseudomonas aeruginosa lectin PA-IIL: 1.0 A resolution crystal structure of the complex combined with thermodynamics and computational chemistry approaches. Proteins. 2005;58(3):735-746. doi:10.1002/prot.20330
- (Mitchell2002) Mitchell E, Houles C, Sudakevitz D, et al. Structural basis for oligosaccharide-mediated adhesion of Pseudomonas aeruginosa in the lungs of cystic fibrosis patients. Nat Struct Biol. 2002;9(12):918-921. doi:10.1038/nsb865
- (Adam2007) Adam J, Pokorná M, Sabin C, Mitchell EP, Imberty A, Wimmerová M. Engineering of PA-IIL lectin from Pseudomonas aeruginosa - Unravelling the role of the specificity loop for sugar preference. BMC Struct Biol. 2007;7:36. Published 2007 Jun 1. doi:10.1186/1472-6807-7-36
- (Reyes1998) Reyes R, Duprat F, Lesage F, et al. Cloning and expression of a novel pH-sensitive two pore domain K+ channel from human kidney. J Biol Chem. 1998;273(47):30863-30869. doi:10.1074/jbc.273.47.30863
- (Zhang2022) Zhang J, Chen J, Lu Y, et al. TWIK-related acid-sensitive K+ channel 2 promotes renal fibrosis by inducing cell-cycle arrest. iScience. 2022;25(12):105620. Published 2022 Nov 17. doi:10.1016/j.isci.2022.105620
- (Li2020) Li B, Rietmeijer RA, Brohawn SG. Structural basis for pH gating of the two-pore domain K+ channel TASK2. Nature. 2020;586(7829):457-462. doi:10.1038/s41586-020-2770-2