Workflow orchestration for genomics

Myelin bridges biological datasets and computational workflows

Discover datasets, structure metadata, generate workflow configurations, and execute reproducible genomics analyses.

Discover datasetsBrowse sample sets

In the browser

Curate before you run

Dataset discovery

Search public datasets, inspect metadata, and curate accessions—the entry point for your workflows.

Open

Sample sets

Structure metadata, contrasts, and batch design into reproducible workflow configurations.

Open

On the cluster

Supported pipelines

YAML workflows in myelin-pipeline-script. Prefer the QC-gated path: preprocess → review in Myelin → differential footprinting. Export inputs from a sample set, then submit with bin/submit.sh.

QC-gated ATAC guide
Default

ATAC preprocessing + QC

Download through peaks and biological QC, then post the report and SCC paths to Myelin for Keep / Exclude review.

DownloadAlign & peaksBio QCQC reportMyelin callback

workflows/atac_preprocessing_qc.yaml

Differential footprinting

After QC review: reuse unchanged merges/peaks when possible, then ATACorrect, footprint scores, and optional BINDetect.

Rebuild if neededATACorrectTOBIASBINDetect

workflows/differential_footprinting.yaml

ATAC footprinting (legacy)

End-to-end without the Myelin QC checkpoint: download through TOBIAS, optional BINDetect, and TF×500 bp bin matrices.

DownloadAlign & peaksATACorrectTOBIASMatrices

workflows/atac_footprinting.yaml

QC for ATAC

Per-sample biological QC, replicate correlation, and aggregated HTML report when BAMs and peak sets are already on disk.

Per-sample QCReplicate correlationQC report

workflows/qc_only.yaml

Public dataset search currently uses NCBI GEO. GEO and related names are service marks of their respective owners—always verify metadata against the original source and publication.