DMT-HAR-MED Skill (Dataset-Orchestration Layer)

Overview

dmt-har-med-skill is the NeuroClaw orchestration skill for the DMT-HAR-MED dataset (OpenNeuro ds006644).

DMT-HAR-MED contains rs-fMRI data from 40 participants in a psychedelic intervention study investigating the effects of N,N-Dimethyltryptamine (DMT) on brain function. The dataset includes multiple experimental conditions (DMT, placebo) and behavioral/physiological measurements.

It coordinates a fixed three-phase workflow:

1. Download DMT-HAR-MED data from OpenNeuro.
2. Prepare and validate BIDS-style data organization.
3. Delegate fMRI processing to fmri-skill.

It also provides phenotype extraction and QC integration paths:

• Extract DMT-HAR-MED phenotype data (intervention conditions, behavioral measures).
• Generate per-subject QC summaries with exclusion lists.

This skill follows NeuroClaw hierarchy:

• Defines WHAT to do, not low-level implementation details.
• Does not execute direct shell commands itself.
• Delegates all execution via claw-shell to base/tool skills.

Research use only.

---

Download Stage (Mandatory First Step)

Source

DMT-HAR-MED data is available on OpenNeuro:

• Website: https://openneuro.org/datasets/ds006644/versions/1.0.1

Supported DMT-HAR-MED Data Packages

• Imaging data: rs-fMRI (NIfTI format)
• Phenotype data: intervention conditions, behavioral and physiological measurements
• Participants: 40 participants with psychedelic intervention

Delegation Rules for Download

• Environment/setup checks: dependency-planner + conda-env-manager
• OpenNeuro dataset download: claw-shell (via openneuro CLI or datalad)
• Optional raw-data organization to BIDS-style staging: bids-organizer

Download Inputs to Confirm in Plan

• Subject list scope (full or custom subset)
• Destination directory with sufficient disk space

---

Core Workflow (Never Bypassed)

1. Identify user target: download, BIDS staging, or full preprocessing.
2. Generate a numbered plan with tools, outputs, runtime, storage, and risks.
3. Wait for explicit confirmation (YES / execute / proceed).
4. On confirmation, run download stage first (if needed).
5. After download success, verify/prepare BIDS staging using scripts/reorganize_dmt_har_med.py.
6. Delegate to fmri-skill for rs-fMRI processing.
7. If phenotype extraction is requested, run scripts/extract_dmt_har_med_phenotype.py.
8. If QC summary is requested, run scripts/dmt_har_med_qc_summary.py.
9. Save outputs into a DMT-HAR-MED-centered structure under dmt_har_med_output/.

---

BIDS Preparation

Script: scripts/reorganize_dmt_har_med.py

Verifies and optionally reorganizes DMT-HAR-MED data to BIDS-compliant layout.

python skills/dmt-har-med-skill/scripts/reorganize_dmt_har_med.py \
  --input /path/to/dmt_har_med_raw \
  --output /path/to/dmt_har_med_bids

Features:

• BIDS validation and verification
• Subject ID normalization
• Modality routing: rs-fMRI
• dataset_description.json and participants.tsv generation
• Dry-run mode: --dry-run to preview

---

Phenotype Extraction

Script: scripts/extract_dmt_har_med_phenotype.py

python skills/dmt-har-med-skill/scripts/extract_dmt_har_med_phenotype.py \
  --phenotype-dir /path/to/dmt_har_med_raw/phenotype \
  --output /path/to/dmt_har_med_output/phenotype/merged_phenotype.csv \
  --imaging-ids /path/to/dmt_har_med_output/bids/participants.tsv

---

QC Integration

Script: scripts/dmt_har_med_qc_summary.py

python skills/dmt-har-med-skill/scripts/dmt_har_med_qc_summary.py \
  --fmriprep-dir /path/to/dmt_har_med_output/fmriprep \
  --output /path/to/dmt_har_med_output/qc/qc_summary.csv \
  --exclude-output /path/to/dmt_har_med_output/qc/exclude_list.csv \
  --fd-threshold 0.3

---

Recommended Output Layout

All assets should be organized under ./dmt_har_med_output/:

• dmt_har_med_output/raw/ (downloaded original files)
• dmt_har_med_output/bids/ (BIDS data)
• dmt_har_med_output/fmri/ (links or copies from fmri_output/)
• dmt_har_med_output/phenotype/ (merged phenotype tables)
• dmt_har_med_output/qc/ (QC summaries and exclusion lists)
• dmt_har_med_output/logs/ (download + orchestration logs)

---

Benchmark Adapter Guidance

For benchmark-style prompts, do not force the full download -> staging -> multimodal processing orchestration when the task is only asking for local DMT-HAR-MED data staging or organization.

• If the task starts from raw DMT-HAR-MED data already present on disk and only asks for BIDS-style staging / validation:
  • skip the mandatory download stage
  • default to the narrow path local raw DMT-HAR-MED discovery -> BIDS validation -> minimal metadata -> report
• In benchmark mode, do not require explicit confirmation before presenting the direct staging solution.

---

Safety and Execution Policy

• No execution before explicit plan confirmation.
• All execution must be routed via claw-shell.
• Missing dependencies must be resolved by dependency-planner before running.

---

Important Notes and Limitations

• DMT-HAR-MED is a specialized psychedelic intervention dataset; analysis requires domain expertise.
• DMT-HAR-MED data from OpenNeuro is already in BIDS format; re-staging may not be needed.
• The dataset includes multiple experimental conditions; event files define condition timing.
• DMT-HAR-MED includes only rs-fMRI; no structural (T1w) or diffusion data is available.
• dmt-har-med-skill is orchestration-only; detailed preprocessing logic remains in fmri-skill.

---

When to Call This Skill

• User asks for end-to-end DMT-HAR-MED workflow.
• User asks to download DMT-HAR-MED data and then run rs-fMRI processing.
• User needs BIDS validation for DMT-HAR-MED data.
• User asks to extract DMT-HAR-MED phenotype data.
• User needs DMT-HAR-MED-specific QC summaries.

---

Complementary / Related Skills

• fmri-skill
• bids-organizer
• fmriprep-tool
• dependency-planner
• conda-env-manager
• claw-shell

---

Reference

• DMT-HAR-MED: https://openneuro.org/datasets/ds006644/versions/1.0.1
• BIDS spec: https://bids.neuroimaging.io/

Created At: 2026-05-06 01:56 HKT Last Updated At: 2026-05-06 01:56 HKT Author: chengwang96