ALLOX Cut GPU Costs by 10x, Halved Support Burden, and Enabled Self-Serve Bioinformatics with Seqera
Read the full case study80-90%
increase in self-service pipeline execution
10x
reduction in GPU costs
50%
reduction in compute support burden
Aim
To build scalable experimental and computational platforms for identifying and characterizing functional pockets in proteins, accelerate the discovery of novel therapeutic targets, and develop the next generation of tools for programmable biology and drug discovery.Challenges
- →Lack of self-serve: Wet-lab scientists depended on the computational team for every dataset analysis, creating bottlenecks.
- →Data integration complexity: Connecting and orchestrating data across tools including Benchling and GCP added significant complexity.
- →Infrastructure and scalability: Managing GPUs, spot instances, and cloud configuration made scaling computationally intensive tasks difficult.
- →Pipeline reproducibility: Continuously evolving lab protocols required constant pipeline adaptation and dataset compatibility.
— André Faure, CTO and Co-Founder
Solution
Nextflow
Building scalable, reproducible pipelines for quality control, processing, and analysis of wet-lab platform outputs.
Seqera
Centralizing pipeline execution at scale, managing testing and production workflows, accessing and sharing results.
Co-Scientist
Accelerating pipeline development through AI-assisted prototyping, debugging, and iteration.
Results
With Seqera, ALLOX achieved a 10x reduction in GPU compute costs, a 50% drop in computational support burden, and enabled their wet-lab teams to self-serve pipeline execution. This has given a small computational team the ability to scale its impact far beyond what headcount alone would allow, eliminating analysis bottlenecks that previously slowed experimental planning, freeing the computational team to focus on building new pipelines and more sophisticated analyses. With their platform throughput expected to double annually, Seqera gives ALLOX the foundation to scale their science and achieve a combination of biological precision and data scale that was simply not possible before.
80-90%
Opportunity
Scalable growth
Enable computational workflows to scale with annual throughput without major investment or restructuring.
Expanded proprietary datasets
Scaling will grow high-quality functional datasets, powering the next generation of predictive models for drug discovery.
Programmable biology
ALLOX's technology has the potential to unlock applications of programmable biology far beyond traditional drug discovery.
About
ALLOX is a biotechnology startup founded by a group of scientists at the Centre for Genomic Regulation (CRG) in Barcelona. Our team envisioned that a powerful new technology we had been developing in the lab could have a transformative impact on the way new drugs are discovered and developed.
That technology is based on Deep Mutational Scanning (DMS) – a technique that enables the engineering of thousands of genetic variants of virtually any human protein and precisely measure their functional consequences on protein-protein interactions directly in living cells. The result is an unprecedented combination of biological precision and data scale that was simply not possible before.
The timing could not be better. Next-generation hit ID and design techniques (including AI chemistry and co-folding methods) can generate binders and identify potential binding sites – yet they remain fundamentally limited: they cannot reliably predict whether a binding event will translate into a functional or therapeutic effect. This is precisely the gap that ALLOX's biophysical data is built to fill. We are generating a wealth of high-quality data for initiating drug discovery programs and training models to predict functional hotspots in silico.
To learn more, visit https://www.allox.bio/