VarSeq runs as either an enterprise server platform accessed through a web browser, or as a standalone desktop application. Choose your deployment below to see the relevant hardware requirements.
VarSeq runs on your server and streams to each analyst's web browser, with nothing to install on their desktops.
VarSeq as a standalone desktop application for individual analysts. VarSeq can also connect to an on-premises VSWarehouse server (the Enterprise Platform) for shared configuration, resources, and licensing.
Enterprise Platform
VSWarehouse, the VarSeq Suite Enterprise Platform, runs VarSeq in web app streaming mode, so users access the full application through their browser. The server uses a Podman-based containerized architecture and runs on Linux. You can connect additional machines, called agents, to run VarSeq sessions and automated analysis workflows as your workload grows.
Size the server by how many analysts run VarSeq sessions at the same time and what they analyze. Each concurrent session needs its own CPU cores and memory on top of the platform's base overhead.
Analysts with a VarSeq session open at the same time, plus any automated workflow runs.
Whole genomes, short read (Illumina) or long read (PacBio HiFi, Oxford Nanopore).
2 users × genome (12 cores + 32 GB each) + 8 cores / 16 GB base platform overhead.
The base platform runs the web portal, user management, and data services. It is already included in the calculator estimate. On its own it leaves no capacity for analysis sessions.
FASTQs, BAMs, VCFs, and project files live on a separate volume from the system SSD. Size it by sample volume and how long you retain raw data. Whole genomes, short or long read, produce the largest files.
NFS/SMB network-attached storage recommended.
Run everything on a single server, or pair a base server with agent servers that host VarSeq sessions and workflow runs. With agents, you split the recommended cores and RAM across machines and can add capacity later as your team or sample volume grows.
All services on one machine
Dedicated agent servers for additional capacity
Golden Helix provides Terraform scripts that provision the server, networking, and storage in your account. Agent VMs start when sessions and workflows need them and scale back down when demand drops.
With permissions to provision VMs, networking, and storage
Choose a region close to your users for low latency
Request a vCPU quota increase in your chosen region for dynamic agent scaling
AWS: On-demand standard instances (8 vCPU base + dynamic agents). Azure: DSv5 (8 vCPU base), BS + FSv2 families (dynamic agents).
Intelligent-tiered cloud storage is configured and mounted to the server for FASTQs, BAMs, VCFs, and other workflow inputs. No upfront limit: you pay only for storage used per month.
Data automatically transitions to cheaper tiers based on access frequency.
Estimate monthly cloud deployment costs based on your planned sample throughput.
Run VarSeq as a standalone application on your own workstation. Hardware needs grow with the size of what you analyze, from gene panels to whole genomes.
When running VarSeq on a shared server with multi-user access (via Remote Desktop or SSH with X11 forwarding), multiply the per-user resource recommendations above by the number of concurrent users. For example, four concurrent exome analysts would need at least 64 GB RAM and 32 cores. Each user should have a separate OS login for a dedicated session.
VarSeq requires outbound HTTP/HTTPS access for annotation downloads and license validation. Proxy server configuration (SOCKS5 or HTTP/HTTPS tunneling) is supported. Organizations with strict firewall policies may need to whitelist Golden Helix server addresses.
Full hardware specifications, OS compatibility details, proxy configuration, and firewall whitelisting addresses.
Our technical team can help you plan hardware specifications, review your infrastructure, or walk through deployment options.