Arcus Biosciences is at the forefront of designing combination therapies, with best-in-class potential, in the pursuit of cures for cancer. By using Nextflow Tower from Seqera Labs, Arcus was able to improve productivity, ensure pipeline traceability, and use cloud resources more efficiently. They were also able to prepare for future growth by scaling capacity for research and clinical trials while providing an intuitive, collaborative user experience to researchers and clinicians.
Arcus is a clinical-stage, global biopharmaceutical company focused on developing differentiated molecules and combination therapies for people with cancer. With a world-class discovery engine and a highly experienced team, Arcus is focused on targeting well-understood biological targets and pathways where there is an opportunity to develop new medicines with a potentially improved safety and efficacy profile. Arcus has multiple investigational medicines in Phase 3 trials and additional investigational candidates at various stages of development in areas of high unmet needs.
To support the evaluation of novel combinations of investigational medicines, Arcus relies heavily on bioinformatics pipelines for gene sequencing and expression, protein quantification, and T-cell profiling. Arcus supports both a pure research environment, as well as an environment supporting clinical trials where the handling of personally identifiable information is a paramount concern.
To scale research and clinical trials, accessibility, ease of use, and the ability to organize pipeline runs and datasets are major issues. The previous pipeline manager used by Arcus was command-line based and difficult for researchers with limited Linux skills to use. As a result, research teams relied on a few skilled individuals to submit and monitor bioinformatic pipelines creating a significant bottleneck. Ideally, Arcus needed a system providing self-service access to researchers with the ability to share results selectively and securely with clinicians and other stakeholders.
Issues such as traceability, repeatability, and auditability are critical. Arcus needed to keep pipeline runs related to specific candidates and clinical trials organized and securely partitioned to meet rigorous documentation requirements.
Another challenge faced by Arcus was using cloud resources efficiently. While the previous pipeline manager could dispatch tasks to the AWS cloud, it was inefficient in terms of resource utilization leading to unnecessarily long runs and overspending.
Arcus began shifting pipelines to Nextflow approximately two years ago to take advantage of high-quality community-curated pipelines and build a more scalable foundation for research and clinical trials.
Shortly after, Arcus deployed Nextflow Tower, a portal for Nextflow workflows designed to enable collaborative data analysis at scale. Tower provided an intuitive interface for researchers and clinicians and supported the notion of workspaces, enabling Arcus to logically partition their research environments and grant access selectively to teams working on various early-stage molecules or those supporting specific clinical trials.
An attractive feature of Tower was that it offered mature integrations with popular cloud services, including AWS Batch, Amazon S3, and Amazon FSx for Lustre. With minimal optimization, pipeline tasks could be packed more efficiently into available cloud resources translating into faster execution and more capacity per unit expenditure in cloud resources.
Tower also insulates researchers and clinicians from underlying compute environments, providing Arcus with the flexibility to run pipelines on AWS, third-party clouds, or on local infrastructure, depending on evolving needs. In addition to their AWS environment, Arcus has since deployed a separate on-prem Slurm-based cluster accessible through the same Nextflow Tower interface providing additional flexibility around where to run pipelines.
Significant cost savings
The combination of Nextflow Tower’s mature cloud integrations and effective resource controls has also helped significantly reduce costs. Arcus estimates that with minimal tuning, the average cost per run decreased by 40-50%. Given their monthly six-figure cloud bill, this is significant. Even more important is that these improvements have provided additional capacity enabling researchers and clinicians to perform additional analysis, improving quality, confidence, and helping accelerate discovery.
A foundation for future growth
By providing users with an intuitive environment supporting early-stage research and clinical trials, Arcus has improved the effectiveness of its research environment and positioned itself for future growth.
By decoupling the research portal from underlying compute environments, Arcus has the flexibility to deploy workloads where it makes the most sense – either tapping scalable cloud resources or taking future workloads in-house.
With Towers’ flexible workspace model, it is easy to create secure collaborative spaces allowing researchers, clinicians, and stakeholders outside the organization to collaborate during clinical trials.