What’s New on Vecura | August 03–07, 2026

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

This week, Vecura expands its collection of computational models and tools with 10 new additions spanning multiple areas of life science research.

Researchers can now access additional capabilities for cyclic peptide and multi-state protein design, antibody structure prediction, single-cell and spatial analysis, genomic quality control, and synthesis planning.

Here is what’s new from August 3 to August 7, 2026.

Models and Tools

Protein and antibody design

Three new additions expand the options available for computational protein and antibody research.

CyclicMPNN

CyclicMPNN is designed for cyclic peptide sequence generation. Fine-tuned from ProteinMPNN, it generates sequences intended to fit cyclic peptide backbones and can support de novo sequence design and motif inpainting.

DynamicMPNN

DynamicMPNN extends protein sequence design to proteins with multiple conformational states, generating sequences compatible with multiple conformations rather than a single static structure.

FlashABB

FlashABB provides fast antibody structure prediction from antibody sequences, supporting structure-based analysis at larger scales.

Together, these additions broaden the protein design capabilities available through Vecura, from cyclic peptides and conformationally dynamic proteins to antibodies.

Single-cell and spatial analysis

Five new tools expand the options available for single-cell and spatial biology workflows.

CopyKAT

CopyKAT uses scRNA-seq data to infer genome-wide aneuploidy, helping distinguish malignant from normal cells and identify tumor subclones.

Harmony

Harmony supports the integration of complex single-cell datasets, helping account for dataset-specific variation across samples, donors, tissues, or technologies.

Scrublet

Scrublet identifies potential doublets in single-cell RNA-seq data, supporting quality control before downstream analysis.

SingleR

SingleR provides reference-based cell type annotation by comparing expression profiles with labeled reference transcriptomic datasets.

FlashDeconv

FlashDeconv supports spatial transcriptomics deconvolution, estimating cell-type proportions from spatial data with an emphasis on efficient analysis at scale.

Together, these tools extend Vecura workflows across data integration, quality control, cell annotation, tumor analysis, and spatial deconvolution.

Genomic quality control

Somalier

Somalier supports sample identity and relatedness checks using sequencing data.

It can be used in workflows for detecting potential sample swaps, assessing relatedness, and performing sample-level quality control across genomic datasets.

Synthesis planning

Selenium Route Search

Selenium Route Search extends Vecura’s capabilities into synthesis planning.

It uses an agentic approach to retrosynthesis to develop, evaluate, and refine potential synthetic strategies for target molecules.

Its addition provides another way to connect computational molecular design with the practical question of how a proposed molecule could be synthesized.

UX and Performance

A smoother Vecura experience

Alongside these new scientific capabilities, a small bug fix was released to improve overall user experience and platform performance.

While this update does not introduce new scientific functionality, it contributes to a smoother day-to-day experience when working on Vecura.

Broader capabilities for computational research

This week’s additions broaden the research workflows available on Vecura, spanning protein and antibody design, single-cell and spatial analysis, genomics, and synthesis planning.

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

Explore Vecura at app.vecura.com.