🟤Data Storage

Data Storage, Collaboration, and HPC Resources at reNEW, UCPH

Data Storage, Collaboration, and HPC Resources at UCPH

Introduction

Biomedical research at reNEW – The Novo Nordisk Foundation Center for Stem Cell Medicine, UCPH, is data-intensive, often involving large imaging datasets, sensitive patient information, and complex analyses. Effective Research Data Management (RDM) ensures data security, meets legal and ethical obligations (including GDPR compliance), and supports high-quality, reproducible science.

A critical component of RDM is data storage management throughout the Research Data Lifecycle. Each stage—from planning to archiving—requires careful consideration of storage infrastructure, data privacy, security protocols, and collaboration needs, especially in projects that involve personal data or multi-institutional teams.

This guide provides an overview of best practices and the University of Copenhagen (UCPH) facilities supporting data storage, sharing, and high-performance computing (HPC) tailored to the needs of biomedical researchers at reNEW.

UCPH Data Storage and Collaboration Solutions

Below is a summary of UCPH facilities suited to various data types and collaboration needs:

Personal Drive

• For individual use (Windows/Linux)

• Suitable for small amounts of non-sensitive research data

• Not designed for sharing with collaborators

Shared Network Drive - Windows / Linux

• For collaboration among UCPH employees

• Ideal for research groups needing shared access

• Not suitable for storing sensitive or personal data

S-Drive (Shared Drive for Sensitive Data)

• Designed for sharing sensitive data among designated UCPH users

• Complies with GDPR

• Suitable for storing and sharing confidential or personal data internally

Microsoft OneDrive for Business

• Cloud storage meeting UCPH security standards for non-sensitive data

• Up to 5TB for sharing with colleagues and external partners

• Covered by a data processing agreement with Microsoft

• Recommended cloud solution (note: Dropbox and other providers do not have such agreements with UCPH)

ERDA (Electronic Research Data Archive)

• Managed by the HPC Center, Faculty of Science

• Centralized storage, archiving, and synchronization service

• Ideal for non-sensitive research data

Sensitive Information Facility (SIF)

• For storing and sharing sensitive information, including personal data

• Enables secure sharing with UCPH and external partners

• Requires GDPR compliance and pre-approval

• Currently provided by the SCIENCE HPC Center; expansion to all UCPH researchers under consideration.

DATA DOI Service

• Built on ERDA

• Data sharing is open to all UCPH employees.

• Facilitates assigning DOIs to datasets for findable, citable research outputs

UCPH Policy for Research Data Management

UCPH Policy for Research Data Management emphasizes:

• Choosing an appropriate, secure infrastructure

• Supporting collaboration while ensuring compliance with information security policies and legal requirements

• Protecting personal data in line with GDPR

Researchers at reNEW must carefully select solutions that meet their data sensitivity and collaboration needs while upholding UCPH policies.

High-Performance Computing (HPC) for Biomedical Research

Biomedical research at reNEW often involves large-scale imaging analyses, computational modeling, and multi-omics integration that require HPC resources. UCPH offers access to both local and national HPC systems tailored to diverse research needs:

Key National HPC Systems Available to UCPH Researchers

Type 1 – Interactive HPC

• Focus on interactivity and easy user access

• Includes YouGene Cluster hosted at UCPH (via UCloud)

• Supports iterative data exploration and analysis

Type 2 – Throughput HPC

• Optimized for high-volume data processing

• Systems include Computerome 2.0, GenomeDK, and Sophia.

• Suitable for large-scale, batch analyses

Type 3 – Large Memory HPC

• Designed for workloads needing large, globally addressable memory

• Hosted at UCPH

• Ideal for complex, memory-intensive biomedical simulations

Type 5 – LUMI Capability HPC

• Pan-European pre-exascale system

• Located in Finland (EuroHPC initiative)

• Suitable for cutting-edge, large-scale computational needs

Special Note: DAN HPC at reNEW

• Owned by DanStem/reNEW, SUND, CGEN/ICMM, CPR at SUND, co-admin with KU

• Shared CPU server for Genomics Data Analysis

• Specialized GPU node for image data analysis

• Connects large datasets stored on KU-IT storage for advanced computation

• See DAN System User Guide for more information

Access and Support

• UCloud provides a user-friendly interface for accessing many of these HPC resources, with guidance and application processes streamlined for researchers.

• Support is available via UCPH IT and the Faculty of Science HPC Center.