Data Sharing of the Phase II from the DIRECT Consortium

Background and Rationale

Major depressive disorder (MDD) is among the top global contributors to disability and disease burden. Despite rapid advances in brain imaging, the translation of neuroimaging biomarkers into clinical psychiatry has been limited—primarily due to inconsistent findings, small sample sizes, and analytic heterogeneity. To address these critical barriers, the Depression Imaging REsearch ConsorTium (DIRECT) was launched in 2017, aiming to build large-scale, harmonized datasets for the scientific community.

The first phase of DIRECT—the REST-meta-MDD Project—aggregated resting-state fMRI data from 1,300 MDD patients and 1,128 healthy controls across 25 cohorts, using a standardized preprocessing pipeline (DPARSF). These data were openly shared (https://rfmri.org/REST-meta-MDD) and have already enabled influential discoveries (see Yan et al., PNAS 2019; Chen et al., Psychoradiology 2022). However, REST-meta-MDD Phase I was limited to region-of-interest (ROI) level indices.

What’s New in DIRECT Phase II

Building on the success and collaborative infrastructure of REST-meta-MDD, DIRECT Phase II has now assembled and harmonized a dramatically expanded dataset—comprising voxel- and vertex-level BOLD time series and T1-weighted structural MRI from 1,660 MDD patients and 1,341 healthy controls across 23 cohorts. All participating sites used the advanced DPABISurf pipeline (see Yan et al., Science Bulletin 2021), and all researchers completed dedicated training to ensure maximal harmonization and quality control.

This resource represents the largest, most detailed open neuroimaging dataset for MDD to date, providing unprecedented opportunities for reproducible, high-powered investigation of functional connectivity (FC) alterations in depression, biomarker discovery, and translational research—from basic mechanisms to individualized intervention strategies (Chen et al., Science Bulletin 2025).

Key Features

• Scale: 1,660 MDD patients + 1,341 healthy controls; 23 cohorts; harmonized protocol
• Data: Voxel/vertex-level resting-state fMRI time series, T1 structural images, standardized phenotypic data
• Pipeline: Unified preprocessing (DPABISurf); all sites trained for consistent data handling
• Community Impact: Enables cross-site reproducibility, large-scale data-driven biomarker discovery, and benchmarking of new analytic methods

Detailed information about each site and imaging parameters can be found in the Supplementary Information of Chen et al. (2025, Science Bulletin).

Data Sharing Process and Requirements

Following the successful data-sharing model established in REST-meta-MDD, DIRECT Phase II data will be shared in two phases:

1. Coordinated Sharing (from publication until Jan 1, 2026):
   • Interested researchers submit a brief letter of intent to rfmrilab@gmail.com.
   • Proposals are reviewed for potential overlap. Approved proposals are notified of ongoing projects, helping researchers avoid duplication and fostering collaboration.
   • We welcome new, innovative proposals—especially those that expand the field or apply novel analytic methods.
2. Unrestricted Sharing (from Jan 1, 2026):
   • Full open access to all qualified researchers.
   • Analyses and applications are welcomed, provided that standard citation guidelines and ethical requirements are followed.

For data access, please visit: https://rfmri.org/DIRECTPhase2

Citation and Acknowledgement Requirements

All users must cite the foundational Science Bulletin paper describing the DIRECT Phase II dataset:

Chen X, Lu B, Wang Y-W, et al. Subgenual anterior cingulate cortex functional connectivity abnormalities in depression: insights from brain imaging big data and precision-guided personalized intervention via transcranial magnetic stimulation. Science Bulletin 2025. https://doi.org/10.1016/j.scib.2025.05.042

We also recommend citing the following related resources to properly acknowledge the consortium’s ongoing work and community standards:

• Chen X, Lu B, Li H-X, et al. The DIRECT consortium and the REST-meta-MDD project: towards neuroimaging biomarkers of major depressive disorder. Psychoradiology 2022;2(1):32-42. https://doi.org/10.1093/psyrad/kkac005
• Yan C-G, Chen X, Li L, et al. Reduced default mode network functional connectivity in patients with recurrent major depressive disorder. PNAS 2019;116(18):9078-83. https://doi.org/10.1073/pnas.1900390116
• Yan C-G, Wang X-D, Lu B. DPABISurf: data processing & analysis for brain imaging on surface. Science Bulletin 2021;66(24):2453-55. https://doi.org/10.1016/j.scib.2021.09.016

Prompt citation and proper acknowledgment support ongoing data sharing and community resource building—please help us make this dataset maximally visible and impactful.

Acknowledgements

This work was supported by the Beijing Nova Program of Science and Technology, the Beijing Natural Science Foundation, the National Natural Science Foundation of China, Sci-Tech Innovation 2030, the National Key R&D Program of China, the Key Research Program of the Chinese Academy of Sciences, and many others (see full list in the Science Bulletin paper).