What’s New on Vecura | July 27–31, 2026

Welcome to What’s New on Vecura, our regular recap of new scientific capabilities and improvements across the platform.

This week’s most significant updates focus on reducing interruptions throughout the research process. Vecura Agent can now help resolve input errors that prevent tools from running, personal and project knowledge is easier to access and manage, and several improvements make workflow runs and reruns more reliable.

Vecura has also expanded its protein–protein interaction capabilities and refined the configuration of AlphaFold3 and Gnina covalent docking workflows.

Here is what’s new from July 27 to July 31, 2026.

Agent

Run tools with Agent-assisted input correction

Invalid or incomplete inputs should not bring an entire research workflow to a dead end.

When an input error prevents a tool from running, users will now see an Ask agent to fix this option. Vecura Agent can analyze the issue, suggest the required corrections, and preview the proposed changes before they are applied directly to the tool configuration.

If information from an uploaded file is required, the Agent will request the user’s consent before accessing it. Once the corrections have been reviewed and approved, the updated inputs can be applied and the workflow can continue without being rebuilt from the beginning.

This creates a smoother tool-running experience while keeping users informed and in control of every proposed change.

Give Vecura Agent personal and project context

Knowledge is now easier to find and manage through a dedicated Knowledge page at /knowledge, rather than being located within general Preferences.

Users can provide Vecura Agent with two types of context:

• Personal Knowledge: Recurring preferences, working conventions, and background information relevant to the individual user

• Project Knowledge: Files, information, and context specific to a particular research project

This additional context helps Vecura Agent provide responses and assistance that are better aligned with the user’s needs and the requirements of each project.

The update makes this existing capability more visible and gives users a clearer place to review and manage the information available to the Agent.

Workflows

Run and rerun workflows with greater reliability

Several improvements make it easier to configure, modify, and rerun computational workflows:

• Workflow validation now updates correctly when a node is deleted, preventing outdated cycle references from blocking the run.

• File-only input ports now capture inline uploads correctly when a workflow run is created.

• Input changes made before rerunning a workflow are applied to the correct fields.

Together, these improvements reduce avoidable errors and make it easier to refine a workflow without rebuilding it from the beginning.

Share workflow outputs more easily

External collaborators can now open outputs from a shared workflow run without signing in.

This makes it easier to share permitted results with research partners, reviewers, and other stakeholders while maintaining a direct connection between the workflow and its outputs.

Models and Tools

Vecura has added two protein–protein interaction scorers to its workflow environment.

DSMBind

DSMBind expands the available options for evaluating interactions between protein partners within broader computational workflows.

PRODIGY

PRODIGY estimates the binding affinity of protein–protein complexes using structural information.

Together, these additions give researchers greater flexibility when evaluating protein complexes or building workflows that require protein–protein interaction scoring.

Refined AlphaFold3 and Gnina configurations

AlphaFold3 now uses conditional assembly inputs. Assembly-related fields are displayed according to the selected configuration, creating a clearer and more relevant tool setup experience.

Gnina covalent docking workflows now support covalent anchor atom and flexible-residue inputs throughout the workflow. The anchor atom defaults to SG and can be adjusted to suit different docking requirements.

These updates provide more precise control over advanced structure prediction and covalent docking workflows.

Resources

Discover content and contributors more easily

Vecura’s redesigned Resources landing page is now live, making it easier to explore publications and learn more about the people behind them.

Blog posts can now feature multiple authors, supported by dedicated author and contributor profiles.

Webinar pages have also been refined so that the Watch recording option only appears when a recording is available.

Supporting more continuous research

This week’s updates help researchers move more smoothly from tool configuration to workflow execution, review, and collaboration.

With Agent-assisted input correction, accessible personal and project knowledge, more reliable workflows, and expanded scientific capabilities, Vecura continues to reduce interruptions across computational research.

Follow What’s New on Vecura for regular updates on new models, scientific workflows, and platform capabilities.