ENIGMA OCD Working Group - Teleconference Minutes

Date: 02/19/2015

Time: 6-7am Pacific Standard Time (3pm Amsterdam)

Re: ENIGMA OCD Updates

Host: Odile A van den Heuvel <

Attendees Present:

~30 people (not 100% about this list, please check)

Odile van den Heuvel (VUmc)

Premika Boedhoe (VUmc)

Dan Stein (Cape Town)

Paul Thompson (UCS)

Neda Jahanshad (UCS)

Lianne Schmaal (VUmc)

Kate Fitzgerald (Michigan)

Carlos Soriano-Mas (Barcelona)

Marcello Hoexter (Sao Paulo)

Yun & Chen (Kwon group, Seoul)

Takashi Nakamae (Kyoto)

Janardhan Reddy (India, but lost connection halfway)

Mick Stevens (Yale)

Francesco Benedetti (Milano)

Patricia Gruner (Yale, bit later due to problems connecting)

Luisa Lazaro + colleague (Barcelona)

Chaim Huyser (Bascule, Amsterdam)

Jan Beucke + colleague (Berlin)

Edna Grunblatt + Silvia Brem (Zurich)

Rachel Marsh (succeeded in connecting?)

Fabrizio Piras (Spalletta group, Rome)

Sarah Madsen (UCS, joined late 6:10am PST, phone mixup)

Noam Soreni (MacMaster, 30 mintes late)

….

Wang (Shanghai)?

Nakao (Fukuoka)?

Cheng (Kunming)?

Paul Arnold (?)

Attendees Not Present:

Derrek Hibar

Carol, Jim Knowles (GWAS)

Guido van Wingen (AMC)

David Mataix-Cols (Karolinska)

Dick Veltman (VUmc)

Invited:

<>

<>

<>

<>

<>

<>

<>

Carles Soriano Mas <>

<>

<>

<>

<>

<>

Chaim Huyser <>

<>

<>

<>

<>

<>

<>

<>

<>

<>

<>

<>

<>

<>

<>

<>

Gruner Patricia <>

cyq <>

Grünblatt Edna <>

Van Erp Theodorus <>

<>

Boedhoe PS <>

Zhen Wang <>

'윤제연' <>

Myongwuk Chon <>

Veltman DJ <>

Lochner C Prof <>

'Poletti Sara' <>

Guido van Wingen () <>

<>

<>

<>

<>

Wit S de <>

Vries Froukje de <>

Werf YD van der <>

<>

<>

<>

<>

<>

<>

<>

<>

<>

<>

<>

<>

<>

<>

<>

<>

<>

<>

'Danielle Cath' <>

Action Items:

  1. Complete subcortical meta-analysis

- Person: all sites (any sites that may not make this deadline should contact Odile, who can offer help in making this deadline)

- Due: 03/31/15 (end of next month)

0.Check your site’s status on the planning sheet, update Odile if status is not correct

- Person- all sites

- Due: Friday 02/27/15 (within one week)

0.If interested in using a method other than the standard FS 5.3 for subcortical segmentations, send Odile some information on your proposed alternative method, including some evidence on validation and/or comparability to the standard FS segmentations

- Person: any site interested in using a method other than the standard FS 5.3

- Due: Friday 03/31/15 (end of next month - this is the proposed deadline for completion of subcortical meta-analysis)

0.Post analysis plan on the ENIGMA OCD website

- Person: Odile (will send info to Neda), Neda/Sarah (update the website)

- Due: as soon as the analyses plan have been submitted to Odile

0.Look at the analysis plan info on the ENIGMA OCD website and email Odile with any ideas for additional analyses

- Person: all sites

- Due: dependent on 4

0.Email Odile to express interest in leading a secondary analysis (ex- in DTI or resting state fMRI)

- Person: all interested sites

- Due: Friday 03/31/15 (end of next month - this is the proposed deadline for completion of subcortical meta-analysis)

Note:

a) prior to the call, the Kwon-group (Seoul) would like to lead an analyses on cortical covariance analyses. > please, fill in analyses plan

b) soon after the call, Odile received the proposal from Fabrizio Piras (Spalletta group), who is willing to coordinate the DTI analyses > additional skype call with Odile soon to discuss the ideas.

0.Contact Carol and Jim for updates on GWAS, update all sites by email

- Person: Odile

- Due: Friday 02/27/15 (within one week)

0.Comment on any issues you want Odile to present at the upcoming ENIGMA PI meetings (Amsterdam Feb 25 and Honolulu June 12)

- Person: all sites

- Due: Friday 02/20/15 (for the Netherlands meetings) or late spring (for pre-OHBM ENIGMA planning meeting)

0.Express interest and time/place preference for an informal meeting of the OCD working group, at upcoming conferences (SOBP Toronto, HBM Honolulu)

- Person: all sites (then Odile will coordinate time/place)

- Due: Friday 02/27/15 (within one week)

0.Attend meeting in Netherlands and update the group on developments and information from the meeting

- Person: Odile van den Heuvel

- Due: Friday 02/17/15 (meeting is next week)

Discussion:

  • See for leading slides the ppt with all information on Quality Check, status of the consortium and additional issues
  • Premika summarized the various steps in the Quality Control (see slides and contact Odile/Premika in case of questions)
  • Poor segmentation pallidum and putamen in FS 5.3. Odile/Premika summarized their attempts to compare FS 5.3, FS 5.0 and FSL First on their own data set (and for FS 5.0 also on Spalletta’s group sample). They will perform additional comparisons on a child OCD sample and Spalletta’s data with First. Discussion on the issue of poor segmentation of these brain regions will be continued during the Feb 25 ENIGMA PI meeting (with Paul Thompson’s group) in Amsterdam.
  • Paul- in one previous ENIGMA study (Stein et al., Nature Genetics, April 15 2012), a small number of cohort PIs were unhappy with the FreeSurfer results
  • ex- Ian Deary- preferred FSL first, which worked slightly better in his elderly cohort (Lothian Birth Cohort, LBC1936, from Edinburgh, UK).
  • so long as a group is happy with the majority of their FreeSurfer segmentations, then they can exclude any that are poor;
  • In ENIGMA1 it was best to allow the sites to decide between Freesurfer and FSL, but in ENIGMA2 (GWAS in 50 cohorts, everyone used Freeesurfer)
  • Odile-
  • first let’s do the meta analysis on FS5.3, excluding those that are very bad
  • then do another analysis on some cohorts using FSL FIRST, to see if any better, and can use the 2 meta-analyses
  • do not like the idea of doing one meta-analysis with two different methods
  • Paul- only a very small handful of groups were unhappy with the FS segmentations, in the ENIGMA1 paper. Jason Stein added some information on FSL vs FS in the supplement of the 2012 paper
  • paper- Stein et al 2012, Nature Genetics
  • Odile- next week in Amsterdam there will be a group meeting, to discuss technical details, after it I will send everyone an email
  • and will do some comparisons and then email everyone how to proceed
  • for now will proceed with FS 5.3
  • Odile- there is also an issue of quality with the segmentation of pediatric scans, which tend to be worse than the adult scan segmentations
  • Kate- segmentations are not working very well with FS, have a colleague at UMich with a large youth population who also had FS problems, this person is working with someone at the NIH to make a child atlas that can be used in conjunction with freesurfer to improve the segmentations
  • question for Odile and Paul- would it be ok to use a different atlas, would the data be comparable?
  • Paul- using different template should be fine so long as you evaluate that the resulting segmentations match anatomy (e.g., expertly/manually defined ground truth), and also evaluate if the customized priors used result in generally larger/smaller segmentations - it would be important to know if the new child atlas led to a difference in the mean size of segmentations, and their variance, relative to the standard Freesurfer method.
  • should be fine for groups to use a different method, as long as the group demonstrates that the method was comparable and valid; valid measures are more important than sticking with a method that gives less accurate results just for the sake of harmonizing the method
  • this is a good suggestion, and this is what other working groups (ex- OCD) have done
  • Odile- current status of whole consortium
  • all sites should look at their own status and see if have correct information listed
  • reminder that overall progress is limited by the slowest group, it is important to move along and get results
  • goal is to do meta-analysis in one month
  • each site should consider if this is possible, if not, then please let us know and we can help figure out how to make it happen (ex- may give some help with doing analysis)
  • hope to do analysis in April, and have results in May, present results in May, and then will write up the paper
  • other ENIGMA working groups (Depression, Schizophrenia, Bipolar)- are ahead and already doing cortical and other secondary analysis, for our group it is important to feel confident in subcortical before moving on, we already have the FS 5.3 processing done, so should be able to publish this soon
  • Odile- analysis plan
  • See draft analyses plan form as used in the other ENIGMA working groups
  • each site can take a look and send info about their ideas
  • the OCD website on ENIGMA is not up yet
  • Neda- please send us the info, and we can put it up, also Odile should have an account to upload their info; we can also offer a specific password protected area just for this group’s members
  • please everyone use the form sent and we will get an overview of the proposed ideas
  • Odile- pooling the raw data for mega-analysis
  • this project has its origin in OBIC, where we pool raw data, so it is worth it to build up a big database of raw data
  • based on informed consent forms from sites, most people are eager to share the raw data (5,000 scans - half cases, half controls)
  • this is about 8 times larger than OBIC (N=780)
  • will enable unique analysis- example clinical phenotype
  • based on experience with OBIC, it is a lot of work to create this large database, needs a lot of QC (ex- for clinical data, missing cells, etc), took about a year to prepare, in the end have a great dataset and people can work in parallel
  • allow different groups to work on different analysis in parallel, worked very well in OBIC
  • Odile will arrange ethical approval at VUmc and sent the English copy of this to the sites.
  • will the coming months to start this, expect to take a long time, once the database is checked and complete can discuss how to divide the work and do specific analysis
  • any comments on building the database for mega-analysis- no comments
  • Odile- so far have used the structural MRI, also have resting state and DTI
  • in the future, can do imaging genetics on the T1
  • if other people have specific interest in RS or DTI, would be great if specific groups can take the lead in some of these -send me an email if interested
  • Odile- the OCD working group can propose specific projects and can seek funding
  • Carlos (Barcelona)- submitted a grant proposal (NARSAD and OCD foundation) for OCD using Resting State data in the ENIGMA groups, with specific interest in the amygdala-frontal cortex connectivity (seed based analysis, amplitude of low frequency fluctuations), the funding would support on RA for 1-2 years in Barcelona and could also help in other analysis in ENIGMA data
  • Odile- asked Carlos if interested in leading a group in OCD on resting state analysis
  • Carlos- can not take the lead, can help
  • Dan Stein & Blair Simpson- put in a grant proposal (NIH) around low income countries (India-Brazil-South Africa), should hear in Q1 of this year, goal is to make neuropsychology batteries more inform. ENIGMA-OCD database is only small (but relevant) part of the grant proposal.
  • everyone is welcome to comment and work with the data. Inform Odile about the plans/submissions in order to keep an overview.
  • Odile- GWAS
  • Carol, Jim (not in the call)- Odile will ask them of the status and let us know by email
  • Odile- options to meet
  • some people from Holland (ADHD, autism, MDD, addiction and OCD working group) and from Paul Thompson’s group will meet next week in Amsterdam
  • another meeting just prior to OHBM with all ENIGMA working group PI’s- Odile will represent the OCD group
  • please let me know beforehand of any issues that you wish to present
  • welcome everyone to send an email with your plans of possible meetings near OHBM (Honolulu) and SOBP (Toronto), host will coordinate and find a common time/place
  • Odile- ask for questions or new ideas
  • no comments/questions
  • Odile- will close the meeting, hope to do the first meta-analysis with one month, if this is a problem at your site, please let us know and we can find how to help you or can do the data processing, thanks and keep up the good work!