RNA-seq workflow with RAPL energy measurement
Workflow
energy, carbon and time per run
One point per workflow run that is used, oldest to newest. Hover for values; click to open the run page. Band and bar: mean ± 1 SD of the runs used, or of each group chosen under “Compare by”.
Runs used for mean ± SD:
Cluster
Nodes
Node use
Run by
Compare by
Exclude runs
untick a run in “values per run” below to leave it out
stage mix per run
One column per run, in chronological order, split by each stage's share of the CPU time of all its stages. The four largest stages have their own colour; the rest are grouped. Hover a column for values; click to open the run.
Overview
purpose
- RNA-seq analysis using FastQC, Salmon transcriptome indexing and quantification, and MultiQC, combined with RAPL CPU-package and DRAM energy measurement during execution. The workflow implementation is nextflow-io/rnaseq-nf. The energy collection method is adapted from CRC-FONDA/RAPL_measurement_workflows by Philipp Thamm and Ulf Leser. Each linked run records its own input dataset, source adaptation, resource limits, traces, and measured results. This combination does not reproduce the original RAPL paper experiment.
FONDA Responsible researcher
- Philipp Thamm Graduate Student
application domain
- Bioinformatics Application domain
Run Metadata
workflow runs
Reproducibility
workflow code link
input datasets used
Execution Environment
workflow engine
- Nextflow Workflow Engine
language
compute cluster
- FONDA Cluster Compute Cluster
backend
- Kubernetes Backend