Quantome: An Interactive Quantum Surrogate Model Explorer

Welcome to Quantome

The Quantome landing page provides a gateway to two distinct modules for exploring how quantum algorithms can be applied to biophysical landscapes:

• Design Your Own Grid: An interactive sandbox where you can build a custom 2D potential energy surface from amino acids and find its quantum ground state using the Variational Quantum Eigensolver (VQE).
• Explore EF-hand Landscapes: A data viewer for exploring pre-calculated VQE results on realistic landscapes derived from protein structures.

---

1. Design Your Own Grid

This module provides an end-to-end pipeline that takes a user-designed classical potential and solves it on a quantum simulator. The workflow is split across three main stages: design, submission, and results.

a) The Design Page

The design page is composed of a control panel on the left and a real-time visualization panel on the right.

The Grid and the Landscape

• Interactive Grid: The 9x9 grid is your canvas. Click on any node to select one of the 20 standard amino acids. Each amino acid is colored based on its charge and its circle radius is proportional to its steric bulk.
• Real-time Preview: As you modify the grid, the 3D surface plot on the right updates instantly, showing you the classical potential energy landscape you are creating.

Control Parameters

You can modify the properties of the landscape using four key parameters:

1. Probe Height ($h_{probe}$): Sets the overall distance of a virtual "probe plane" from the amino acids.
2. Minimum Probe Distance ($\epsilon$): Defines a steric "keep-out" zone. If the Probe Height is too low for a bulky amino acid, a "Steric Clash" will be detected.
3. Probe Charge ($q_{probe}$): Sets the charge of the virtual probe to be either +1 or -1.
4. Hopping Strength ($t_{hop}$): A quantum-mechanical parameter that controls tunneling between grid sites.

Sanity Check and Submission

• Calculate Ground State: This button performs an instant, on-page calculation to find the minimum of the classical potential energy, providing a quick diagnostic.
• Train VQE and Sample Probabilities: This button packages your grid design and all parameters and sends them to the backend for a full VQE simulation.

b) Job Submission & c) Results Page

After submission, you are redirected to a status page with a unique Job ID. Once the job is finished, you are automatically taken to the results page, which provides a comprehensive visual summary of your simulation.

---

2. Explore EF-hand Landscapes

This module allows you to explore the results from applying the surrogate model pipeline to a set of real-world biological systems: calcium-binding EF-hand loops. Unlike the design module where you create the landscape, here you can inspect pre-calculated results for these protein binding sites.

a) The Selection Page

You are first presented with a table summarizing the pre-calculated protein systems available for exploration. Each row corresponds to a specific protein chain and provides basic structural information. Click on the "View Results" link for any entry to proceed to the detailed analysis page.

b) The Detailed Results Page

This page provides a comprehensive breakdown of the VQE simulation for a specific protein.

Ion Site Selection & 3D View

If a protein contains multiple Ca²⁺ binding sites, a dropdown menu allows you to select the specific site you wish to inspect. An interactive 3D view of the protein binding pocket, rendered with Mol*, shows the local atomic environment of the selected ion.

Data Visualization

Upon selecting a site, the page loads and displays several key results:

• Potential Energy Surfaces: A series of 3D surface plots show the step-by-step construction of the potential landscape:

  1. Height Map: Represents the steric boundaries of the pocket derived from ray-casting.
  2. Charge Map: Shows the electrostatic features based on the charges of the pocket-lining residues.
  3. Combined Potential: The final classical landscape used as the input for the simulation.
  4. Polynomial Fit: The smooth surface generated by the 10th-degree polynomial fit.

• VQE Probability Heatmaps: Side-by-side 9x9 heatmaps display the final VQE probability distributions. These plots show the most likely location for the ion in the two cases case ($t_{hop}=0$ and for the user defined value), allowing for a direct comparison. The color represents the Enrichment Factor, and hovering over any cell reveals the precise probability and enrichment values.

• Diagnostics: A panel displays key diagnostic values, including the Root Mean Square Error (RMSE) of the polynomial fit and the minimum energy ($E_{min}$) found by VQE for each simulation.