Seven Bridges knowledge centerDocumentation
Release notes
Back to All

Release Note - Recently Published Apps - May 15, 2026

1 day ago by Rucha Kulkarni

MONAI is an open-source framework for deep learning in healthcare. In this release, following 2 applications wrapped with Monai wrapper are added in Public Apps Gallery.

  1. MONAI Auto3DSeg is a wrapper for the AutoRunner utility of Auto3DSeg. It is used to automatically train, tune hyperparameters, and evaluate multiple 3D segmentation models with minimal user intervention.
  2. MONAI nnUNetV2 is a wrapper for the nnUNetV2Runner utility of nnUNetV2. It is a fully automated, self‑configuring deep learning framework for biomedical image segmentation which can automatically train, tune hyper parameters and evaluate multiple 3D segmentation models end-to-end without human tuning by automatically adapting itself to any new dataset.

MOFA2 is an unsupervised multi-omics integration framework that learns latent factors capturing shared and view-specific sources of variation across multiple omics datasets. In this release, following 3 tools are added in Public Apps Gallery

  1. Data Harmonizer reads omics data files (tabular, .h5ad, or .h5mu) and prepares matrices for MOFA2 input structure. It supports matrix and long-format tabular data, along with AnnData and MuData objects, and performs format harmonization, basic data-type-aware transformation, and sample alignment when needed. The app is intended for MOFA2 input preparation and does not substitute for full assay-specific preprocessing or quality control.
  2. MOFA2 takes a set of multi-omics data files and performs Multi-Omics Factor Analysis (MOFA2) to infer latent factors that represent the underlying biological signals across diverse modalities. This process allows for the integration of data types such as transcriptomics, methylation, and proteomics into a unified latent variable model.
  3. MOFAx-0-3-7 generates visualization plots from trained MOFA2 multi-omics models. It helps interpret latent factors and explore relationships among samples, features, views, and covariates derived from multi-omics datasets.

pgsc_calc is a bioinformatics best-practice analysis pipeline for calculating polygenic risk scores on samples with imputed genotypes using existing scoring files. This nextflow workflow is available in Public Apps Gallery on BDC and Igor Production platforms

Resources

If you need help with accessing controlled study details, please contact your Velsera Seven Bridges representative or email [email protected].