Depression Biotypes Multisite rsfMRI (Drysdale 2017)
Multisite resting-state fMRI sample of the depression biotypes study (Weill Cornell, Drysdale et al. 2017)
Resting-state fMRI and structural brain MRI of adults with major depression and healthy controls, pooled to define four connectivity subtypes of depression. 458 patients and 730 controls from 17 sites. Patient sites share data at the investigators' discretion.
Overview
This entry describes the pooled multisite sample behind the depression biotypes study of Drysdale, Liston and colleagues at Weill Cornell Medical College (Nature Medicine, 2017). The authors clustered resting-state functional connectivity of patients with major depression into four subtypes and trained classifiers to assign them. The paper treats its scans as one large multisite data set, split into a training set and an independent replication set. The sites carry labels such as Cornell 1, Cornell 2, Toronto and Stanford 1.
Composition
The sample has 1,188 adults. Data set 1 (training) holds 333 patients in a current major depressive episode and 378 healthy controls from 12 sites. Data set 2 (replication) holds 125 patients (109 unipolar, 16 bipolar II) and 352 controls from 13 sites, five of them new. Supplementary Table 3 lists 17 site labels in total. Cornell 1 contributed 96 patients and 28 controls to data set 1 and 35 patients and 35 controls to data set 2. Cornell 2 contributed 27 patients and 8 controls to data set 2. Clustering used 220 patients from Cornell 1 and Toronto, two sites with the same inclusion criteria (treatment-resistant unipolar depression). Fifty Cornell 1 patients had a second scan 4 to 5 weeks later, and 124 Toronto patients had imaging before a course of rTMS to the dorsomedial prefrontal cortex. The study also used 39 patients with generalized anxiety disorder (Cornell 1 and Stanford 1) and 41 with schizophrenia (COBRE), which are not counted in the 1,188.
Acquisition
Each subject had a resting-state BOLD scan (spiral in-out or Z-SAGA) and a T1-weighted MP-RAGE or SPGR scan. Parameters differ by site; most used a TR near 2 s, about 3.5 mm in-plane resolution and 150 to 180 volumes. Cornell 1 and Toronto scanned on 3T GE Signa systems with 180 and 300 volumes. Depression severity was rated with the 17-item Hamilton scale.
Annotations
No image annotations. Labels are diagnosis, Hamilton item scores, biotype assignment and, for the Toronto rTMS group, treatment response.
Known limitations
- No public release of the patient sites. Sharing is left to each site investigator, with no published terms.
- Controls from ten sites come from the 1000 Functional Connectomes Project and are tracked there.
- Inclusion criteria, medication use and comorbidities differ by site.
- Sex is reported only as percentages per group and site, and scanner models only for Cornell 1 and Toronto.
Cohort
Aggregate numbers from the sources below. Bars are relative to the 1,188 subjects.
Modality
- MRI 1,188 100%
Online Methods: MRI data acquisition in Drysdale et al. 2017, Nature Medicine (author manuscript PMC5624035)
Condition
Groups can overlap
- Healthy control 730 61%
- Major depressive disorder 458 39%
Online Methods: Subjects in Drysdale et al. 2017, Nature Medicine (author manuscript PMC5624035)
Field strength
- 3 T 220 19%
Supplementary Table 1 in Drysdale et al. 2017, Nature Medicine, Supplementary Information
Condition by split
Reported cross table. A dot marks a cell the source does not give.
| Training | Test | |
|---|---|---|
| Healthy control | 378 | 352 |
| Major depressive disorder | 333 | 125 |
Online Methods: Subjects in Drysdale et al. 2017, Nature Medicine (author manuscript PMC5624035)
Contrast / sequence
Groups can overlap
- BOLD fMRI 1,188 100%
- T1-weighted 1,188 100%
Online Methods: MRI data acquisition in Drysdale et al. 2017, Nature Medicine (author manuscript PMC5624035)
Scanner vendor
- GE HealthCare 220 19%
Supplementary Table 1 in Drysdale et al. 2017, Nature Medicine, Supplementary Information
Split
- Training 711 60%
- Test 477 40%
Online Methods: Subjects in Drysdale et al. 2017, Nature Medicine (author manuscript PMC5624035)
Other reported numbers 4
| Mean age · Major depressive disorder, Training | 40.6 | Online Methods: Subjects in Drysdale et al. 2017, Nature Medicine (author manuscript PMC5624035) |
| Mean age · Healthy control, Training | 38 | Online Methods: Subjects in Drysdale et al. 2017, Nature Medicine (author manuscript PMC5624035) |
| Mean age · Major depressive disorder, Test | 49.8 | Supplementary Table 3 in Drysdale et al. 2017, Nature Medicine, Supplementary Information |
| Mean age · Healthy control, Test | 32.1 | Supplementary Table 3 in Drysdale et al. 2017, Nature Medicine, Supplementary Information |
License and access
Our reading of the license, not legal advice. Before you use the data, read the original license and confirm that your use is allowed. We take no responsibility for how you use a dataset. Full disclaimer
Ask the authors or the data holder. No published process or terms, and access is at their discretion
The data availability statement says that data from the sites outside the 1000 Functional Connectomes Project are available at the discretion of the site principal investigators named in Supplementary Table 2 (Cornell 1: Dubin and Liston; Toronto: Downar; Emory 1: Mayberg; Stanford 1: Etkin; Stanford 2: Schatzberg). Control and COBRE data from NKI, Atlanta, Cambridge, Cleveland, ICBM, New York, COBRE, Beijing, Milwaukee and Leipzig are public through the 1000 Functional Connectomes Project.
Different parts of the data carry different licenses. The summary on the right shows the most restrictive answer per rule.
No published license or data use terms
The data holder has published no license and no data use terms. Any use, sharing or commercial right has to be agreed with the data holder, and nothing can be assumed to be allowed. Default copyright and data protection law still apply.
The paper's data availability statement leaves sharing to the site investigators and states no terms.
What you can do
- Commercial use Not stated
- Not stated
- Train ML models Not stated
- Create derived data Not stated
- Publish results Not stated
What you can share
- Share the data Not stated
- Share derived data Not stated
- Share trained models Not stated
What you must do
- Cite or credit Not stated
- Share alike Not stated
- Sign an agreement Not stated
- Ethics approval Not stated
- Manuscript review Not stated
- Release code Not stated
- Return results Not stated
- Delete after use Not stated
Limits
- No re-identification Not stated
- Location limits Not stated
1000 Functional Connectomes Project / INDI terms (Creative Commons Attribution Non-Commercial, no version stated)
Non-commercial use only, after registering with the 1000 Functional Connectomes Project on NITRC. The NITRC project lists a Creative Commons Attribution Non-Commercial license without a version: derived works must credit the source and stay non-commercial but need not use the same license.
What you can do
- No
- Not stated
- Not stated
- Conditional
- Not stated
What you can share
- Not stated
- Conditional
- Share trained models Not stated
What you must do
- Yes
- No
- Conditional
- Ethics approval No
- Manuscript review No
- Release code No
- Return results No
- Delete after use No
Limits
- No
- Location limits No
Commercial license: Not stated
Citation
Drysdale AT, Grosenick L, Downar J, Dunlop K, Mansouri F, Meng Y, Fetcho RN, Zebley B, Oathes DJ, Etkin A, Schatzberg AF, Sudheimer K, Keller J, Mayberg HS, Gunning FM, Alexopoulos GS, Fox MD, Pascual-Leone A, Voss HU, Casey BJ, Dubin MJ, Liston C. Resting-state connectivity biomarkers define neurophysiological subtypes of depression. Nature Medicine 23(1):28-38 (2017). doi:10.1038/nm.4246
Sources
Every number on this page comes from one of these documents. Each chart names the table or page it is taken from. The raw numbers are in stats.csv.