VTX: real-time high-performance molecular structure and dynamics visualization software

Abstract

Summary

VTX is a molecular visualization software capable to handle most molecular structures and dynamics trajectories file formats. It features a real-time high-performance molecular graphics engine, based on modern OpenGL, optimized for the visualization of massive molecular systems and molecular dynamics trajectories. VTX includes multiple interactive camera and user interaction features, notably free-fly navigation and a fully modular graphical user interface designed for increased usability. It allows the production of high-resolution images for presentations and posters with custom background. VTX design is focused on performance and usability for research, teaching, and educative purposes.

Availability and implementation

VTX is open source and free for non-commercial use. Builds for Windows and Ubuntu Linux are available at http://vtx.drugdesign.fr. The source code is available at https://github.com/VTX-Molecular-Visualization.

1 Introduction

Molecular visualization is a critical task usually performed by structural biologists and bioinformaticians to aid the different processes that are essential to understand structural molecular biology (Olson 2018). Following the recent advances in the determination of atomic resolution molecular structures and assemblies (Jumper et al. 2021), in protein structure prediction (Yip et al. 2020, Varadi et al. 2022), and in the increased accessibility of molecular dynamics simulation (Amaro et al. 2024, Hospital and Orozco 2024), there is a profusion of molecular structural biology data that is available to the scientific community. Due to the modern High-Performance Computing hardware and storage, the size of the simulated systems (Jung et al. 2019, Stevens et al. 2023, Coshic et al. 2024) with molecular dynamics has dramatically increased and the storage, analysis and interactive visualization of the resulting data can become problematic for currently available molecular visualization systems such as PyMOL (Schrödinger and DeLano 2020), VMD (Humphrey et al. 1996), or Chimera-X (Pettersen et al. 2021).

Here, we present VTX, an open-source molecular visualization software. VTX is optimized to handle efficiently and in real-time the big data in molecular simulations notably by including a meshless high performance molecular graphics engine coupled with a minimalistic task-oriented graphical user interface (GUI) to maximize the usability for nonexpert users. VTX uses the chemfiles library (Fraux 2021) to read and write molecular data which handles most widely used molecular structures and trajectories file formats. VTX includes various representations and rendering options and provides different tools such as structural alignment and distance/angle measurement for interactive analysis of massive molecular scenes. It is free and open source for non-commercial use and available on linux and windows at http://vtx.drugdesign.fr and https://github.com/VTX-Molecular-Visualization

2 Methods

2.1 Molecular graphics engine

Thanks to adapted data structures and rendering algorithms, VTX is designed for the real-time visualization of very large molecular systems, composed of several millions atoms, on a consumer laptop. A molecule is defined by the position of its atoms, their corresponding radius and connectivity.

2.1.1 Meshless representations

Most of VTX representations are described implicitly, without using a triangular mesh (sticks, ball and sticks, Van der Waals, Solvent Accessible Surface). These meshless representations allow impostor-based techniques in the rendering engine (Bagur et al. 2012). For each primitive, a simple quad is rasterized. Then, ray-casting is used to evaluate the implicit equation of the primitive and display the final shape. This allows fast and pixel perfect rendering while reducing the memory consumption and bandwidth usage which is essential to handle large molecular structure or dynamics data.

2.1.2 Cartoon and SES representations

The Cartoon representation is generated on-the-fly and displayed with adaptive level-of-detail (LOD) method using tessellation shaders (Hermosilla et al. 2015).The Solvent Excluded Surface is computed following a discrete approach, similarly to (Lindow et al. 2014) and the resulting surface is extracted via marching cubes (Lorensen and Cline 1987).

2.1.3 High-quality rendering

VTX aims to enhance the user’s visual analysis experience in real-time by providing high-quality rendering (Fig. 1B). The use of meshless representations in VTX results in a pixel-perfect display quality. Additionally, VTX offers various rendering options to improve the perception of details and enable the creation of visually appealing illustrations. The deferred rendering approach allows for the implementation of various post-processing techniques, including shading (flat, matte, glossy, toon), ambient occlusion, fog, outline, and anti-aliasing. These techniques can enhance the visual perception of molecular shape and improve overall image quality. VTX offers an image export feature that allows to produce an image of the displayed molecular scene with a resolution up to 8K and a custom background with user-defined transparency. When dealing with dynamic data (such as MD trajectories), ambient occlusion rendering cannot be pre-computed. In VTX, we use a deferred shading graphics pipeline to compute lighting on surfaces and apply screen-space post-processes that allows to produce high-quality rendering in real time (Mittring 2007).

Figure 1.

(A) Illustration of the VTX GUI organized with different dockable panels. The Scene Tree panel contains and allows the selection of all the objects present in the scene, such as molecules, labels, and viewpoints. The Sequence panel displays and allows the selection of the sequence of each biomolecule, while the Visualization Panel allows the selection, observation, and manipulation of their 3D structure. Frequently used options for manipulation and rendering are directly accessible in the visualization panel through the use of button overlays. When an object is selected in the scene, detailed information about the loaded object, additionally loaded trajectories and chosen representations are available in the Inspector panel. The Console panel displays all logs. (B) Different molecular representations available in VTX (Saccharomyces cerevisiae BCP1, PDB id: 7C4H) Cartoon (upper left), Van der Waals (upper right), balls and sticks (bottom left), Solvent Excluded Surface (bottom right). (C) Illustration of the coarse-grained whole mycoplasma cell model from (Stevens et al. 2023) with (left) and without (right) clipping. Martini beads are displayed as surface and flat color (membrane lipids in green, membrane proteins in blue, ribosomes in yellow, chromosome in light blue, cytoplasmic proteins in dark blue).

2.2 Improved user experience

2.2.1 Camera modes

VTX disposes of an interactive camera system controllable via the keyboard and/or mouse that includes different modes: (i) trackball and (ii) free-fly. The trackball mode is the classical camera available in molecular visualization software where the camera revolves around a fixed focus point. This allows the user to rotate the view around the object of interest. The first-person free-fly navigation mode allows the user to fully control the movement of the camera. This mode is similar to the first-person perspective in video games where the user can move freely in the 3D space. It has notably been used in UDock (Levieux et al. 2014, Plateau-Holleville et al. 2022) and in the multiscale molecular rendering tool CellVIEW (Le Muzic et al. 2015)

2.2.2 Keyboard controls

Using a trackpad to navigate in 3D molecular scenes can be challenging because it may not provide the same level of precision and control as a mouse. This can make it difficult to perform precise tasks such as rotating, zooming, and panning the view. VTX provides alternative navigation controls using keyboard bindings to make it easier to use with a trackpad and to complement the mouse for the freefly navigation mode.

2.2.3 Customizable GUI

The GUI of VTX enables to create customizable presets of combined displays and representations that can be saved and easily accessed. The different dockable windows of the software can be moved, resized, and arranged according to the user’s preferences, allowing for a customizable workspace. It also includes quick access to the most frequently used commands in the 3D visualization window through the use of clearly labeled buttons.

2.3 File formats compatibility

The Chemfiles library (Fraux 2021) is used to handle diverse file formats for molecular structures (including PDB, mol2, mmCIF and MMTF) and molecular dynamics trajectories (including dcd, xtc, and arc). While mmCIF is the preferred format in VTX for molecular data as its structure is well adapted for very large molecular structures and assemblies, a direct download access to the PDB (Berman et al. 2000) is provided through the use of the PDB API.

2.4 Implementation

VTX is written in C++, with modern OpenGL for hardware-accelerated rendering and the Qt (Nokia Corp. 2012) framework for its GUI.

2.5 Operation

VTX operates on computers supporting OpenGL 4.5. The software is compatible with Windows 10 and subsequent versions, as well as Ubuntu Linux versions 20 and higher.

3 Performance evaluation

The performance of VTX version 0.4.4 in terms of system file reading and fluidity of manipulation has been comparatively evaluated to VMD version 1.9.4 (Humphrey et al. 1996), PyMOL version 3.1 (Schrödinger and DeLano 2020) and ChimeraX version 1.9 (Pettersen et al. 2021). The performance of each software was assessed using a benchmarking dataset comprising diverse molecular systems, characterized by varying sizes ranging from 100 000 atoms to 100 million Martini beads. The evaluation focused on two primary metrics: (i) system stability, defined as the software’s ability to load the molecular system without crashing; and (ii) interactive fluidity, quantified by the smoothness and responsiveness of user interaction during selection/change of representation and manipulation tasks, graded using different levels (−crashed or frozen, +low fluidity, ++moderate fluidity, and +++high fluidity).

The benchmarking dataset is comprised of four different systems: (i) the structure of type IVa pilus machine (Chang et al. 2016) containing 107 640 atoms (PDB ID: 3JC8); (ii) the structure of the P68 bacteriophage (Hrebík et al. 2019) containing 1 074 183 atoms (PDB ID: 6Q3G); (iii) a concatenated model involving three structures: two microtubules (Walton et al. 2023) (PDB ID: 8J07 and 8GLV) and a reconstruction of a bacteroides phage crAss001 (Bayfield et al. 2023) (PDB ID: 8CKB) for a total of 11 003 162 atoms; and (iv) the 2023 Martini minimal whole cell model from (Stevens et al. 2023) displayed in Fig. 1C that contains 60 887 soluble proteins, 2200 membrane proteins, 503 ribosomes, a single 500 kbp circular dsDNA, 1.3 million lipids, 1.7 million metabolites and 14 million ions for a total of 101 753 154 Martini beads. All evaluations have been performed on a Dell Alienware 15r with an intel i7-10750H CPU and a RTX2080 Super. Resulting performance is displayed in Table 1.

Table 1.

Comparative performance evaluation of reference molecular visualization software with VTX on different molecular systems with sizes ranging from 100K atoms to 100 million Martini beads.^a

	System stability				Interactive fluidity
	ChimeraX	PyMOL	VMD	VTX	ChimeraX	PyMOL	VMD	VTX
3jc8 (107 640 atoms)	✓	✓	✓	✓	+++	+++	+++	+++
6q3g (1 074 183 atoms)	✓	✓	✓	✓	+++	+++	++	+++
8ckb 8j07 8glv (11 003 162 atoms)	✓	✓	✓	✓	+	+++	−	+++
Whole cell model (101 753 154 Martini beads)	X	X	✓	✓	−	−	−	++

^aSystem stability measures the ability of the software to load the given system without a crash (✓ successful loading, X crashed). Interactive stability measures the smoothness and responsiveness of user interaction during selection, change of representation and manipulation tasks, graded using different levels (- crashed or frozen, + low fluidity, ++ moderate fluidity, and +++ high fluidity). All evaluations have been performed on a Dell Alienware 15r with an intel i7-10750H CPU and a RTX2080 Super.

No values could be obtained with ChimeraX and Pymol on the 100M system, due respectively to a crash and a freeze during system loading. Similarly, even though VMD could open 10M and 100M systems, the selection and manipulation of the system resulted in a freeze. Despite limited computational resource, VTX allowed to easily manipulate the larger systems in real-time, perform precise selections and modify rendering settings. The free-fly real-time navigation video presented in Supplementary Material was captured with microsoft XBOX overlay on windows 11.

4 Use case

The VTX GUI is organized as presented in Fig. 1A. Molecular structure and trajectories can be loaded using the open button from the file menu or directly downloaded from the PDB with their accession number. For ease of use, molecular trajectories can be loaded with a right click onto an already loaded molecular object in the scene tree.

As illustrated in Fig. 1C, due to its new generation molecular graphics engine, VTX is highly scalable and allows the rendering in real time of massive molecular systems (100+ million Martini beads) on a consumer laptop with a NVIDIA RTX2080m.

VTX also includes different functions in the tools menu tab for distance and angle measurements, as well as structural alignment using the CE method (Shindyalov and Bourne 1998). The session state, including the user’s organization of the VTX workspace, can be saved and exported. High resolution illustrations with or without background can be generated with the snapshot function.

5 Conclusion and perspectives

VTX is a molecular visualization software designed to provide: (i) a high-quality real time rendering of large molecular systems (several hundred million atoms on a consumer laptop), (ii) a comfortable user experience with intuitive controls and tools, and (iii) a wide compatibility with most molecular structures and trajectories file formats and a high-quality image export feature allowing to produce up to 8K-resolution images with custom background for posters and presentations. Future versions will include real-time analytical SES (Plateau-Holleville et al. 2025), high-end real-time ambient occlusion, offline ray-tracer for high-quality illustrations and movies rendering, python-like command binding. VTX is open source and free for non-commercial use.

Author contributions

Nicolas Dacquay (Software [equal]), Cyprien Plateau-Holleville (Software [equal]), Valentin Guillaume (Software [supporting], Writing—review & editing [Supporting]), Vincent Larroque (Software [supporting]), Jean Larde (Software [supporting]), Yassine Naimi (Software [supporting]), and Jean-Philip Piquemal (Writing—review & editing [supporting])

Supplementary data

Supplementary data are available at Bioinformatics online.

Conflict of interest: None declared.

Funding

VTX has received funding from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation program [Grant Agreement No. 640283] and from the ERC under the Horizon Europe research and innovation program [Grant Agreement No. 101069190]. C.P.H. and V.L. are supported by institutional grants from the National Research Agency under the Investments for the future program with the reference ANR-18-EURE-0017 TACTIC. V.L. is a recipient of a fellowship from Qubit Pharmaceuticals.

Data availability

All files composing the benchmarking dataset, except the Martini minimal whole cell model, are available at https://doi.org/10.5281/zenodo.14962673. All .gro files that constitute the Martini minimal whole cell model are available upon request at the Marrink lab. The procedure to generate the model is available at https://github.com/marrink-lab/Martini_Minimal_Cell

Software availability

VTX is open source and free for non-commercial use under the VTX consortium license. Builds for Windows and Linux are available at http://vtx.drugdesign.fr. The source code is available at https://github.com/VTX-Molecular-Visualization.