Scaling Gene Therapy Research with Argo Workflows and Hera
Talkat ArgoCon Europe 2023 · on Tuesday, 2023-04-18· at 14:15· Elicium Building | Elicium Ballroom 2 [Argo Workflows]·
summary
Dyno Therapeutics uses Argo Workflows and Hera to generate and test AAV variants from 750‑base sequences (search space ~20^750), running design jobs with ~500 pods per workflow and 10–15 concurrent workflows, fan‑out/fan‑in patterns, and model training on nodes with ~4 GPUs each. Hera is a Python SDK that replaces YAML with script templates for faster experiments. Pipelines chunk configs to avoid etcd/annotation limits, process 50GB–TB sequencing files in parallel, and perform full rollback on failed ingestion steps.
AI-written summary · about
chapters (7)
- 0:00Start
- 1:21Gene therapy
- 2:29Solving the delivery problem
- 5:37Sequence design
- 6:44Model training
- 8:50When things become a reality
- 13:27Hera: easy access to Argo
related talks
25mPanel: Argo for ML: Achieving Scalability and User Experience
25mOne Engine To Rule Them All: Unifying Cloud Workloads With Argo Workflows
23mOrchestrating Rocket Simulations W/ Argo
27mEmpowering Autonomy: BYD's Journey Taming Million-Task Scale With Argo...
10mScaling Time Series Analysis With Argo Workflows: Patterns and Practi...
25mHow to Train an LLM with Argo Workflows and Hera
25mHow to Multiply the Power of Argo Projects By Using Them Together
26mLeveraging Argo Ecosystem for Scalable Cloud-Native Solutions at Databricks