Frequently Asked Questions (FAQ) about the PGC, secondary analysis proposals, and data access

FAQ - About the PGC

The first way is to join a group that is already part of the PGC at your institution or in your professional network. The second way is to share GWAS data on cases and controls for one of the PGC disorders. Each group makes its own determination on the minimum number of samples an individual or organization can contribute. The best approach is to contact the working group chair to find out the group’s policies about sample size contributions.
The best way is for you join a group at your institution or in your professional network that is already part of the PGC.
Contact the working group chair representing the disorder of the case samples you have to contribute.
Sometimes yes and sometimes no. The best place to start is to contact the chair of the PGC working group because that person will have up to date knowledge about what the possibilities for genotyping are. The PGC mostly funds analysis rather than genotyping.
After a person becomes part of a PGC working group, the investigator gets back the cleaned and imputed version of his or her own data, the results file for his or her data, and usually genetic risk scores computed from all other case-control samples excluding his or her own sample. How these data can be used, guidelines for submissions of secondary analysis proposals and manuscripts are all governed by individual PGC working groups.
Once you become a PGC member, you can join the calls for the working groups to which you belong. Optionally, people can become part of the cross disorder group, the CNV group, the statistical analysis group, and the pathway/network group should these be aligned with their interests. The PGC also has a monthly Worldwide Lab meeting call to which all PGC members are welcome.
Fantastic question, there is always more to do and we always need more help! The best first step is to talk with your working group chair to see what needs to be done. Since we are a voluntary organization, people often get involved by volunteering to do sometimes inelegant and low level tasks. The PGC is a labor of love and we welcome enthusiastic and dedicated workers. We live by the rules of sweat equity and you are rewarded for the effort you dedicate to the consortium. We anticipate turnover of PGC leadership over the next several years, so getting involved now will position you well for the future.
The Psych Chip is an Illumina SNP array containing a common variant backbone, rarer exome content, and 50,000 custom probes designed by the PGC to capture relevant psychiatric content.

FAQ - About Proposals

Before writing your proposal, it is a good idea to check the list of approved and current proposals to make sure no one has already proposed the same analysis. If you find an analysis that closely mirrors your own, we suggest you contact the proposal PI and ask to join the project.
The PGC Data Permission Sheet (DPS) lists data sets by disease group and includes PI contact information, required permission, and sample sizes of cases and controls. Consult the appropriate PGC DAC representative for more information.
There is a secondary analysis proposal form that must be submitted for your project, which details the specific aspects of the project that should be communicated to the workgroup. The project description does not need to be at the level of a grant submission, but you should include enough detail so that the study design and feasibility can be easily understood. Given that each workgroup is comprised of investigators from diverse backgrounds, it is suggested that you avoid technical language or field specific jargon.
Please list, by name, any individuals who will be involved in the research. This includes other investigators (both inside and outside the PGC) who will be involved in the planning, interpretation, and communication of results. This also includes analysts, postdocs, and students who will be performing the actual analysis. If you plan to hire someone to conduct analyses, please write that person’s position in as well with a note that says “to be determined” (e.g. Postdoc - TBD).
PGC PIs are defined as investigators who have shared individual level genotype data with the PGC and are otherwise PGC members in good standing. If you are a PGC PI, you do not need an additional sponsor. If you are not a PGC PI (examples may include graduate students, postdocs, or faculty contributing to the PGC through their analytic talents) your proposal will need to be sponsored by a PGC PI. The sponsoring PI must be directly involved in the work proposed. This rule is intended to encourage collaboration and to reduce barriers to data sharing. It is important to note that the PGC is not a data repository, it is a collaborative space for clinicians, statisticians, and geneticists to exchange both ideas and data. PI sponsorship is both an incentive to data sharing and a way of increasing interactions between investigators of different backgrounds who are interested in a particular research question.
Yes. Analysts from outside of the PGC are welcomed to submit proposals. Proposals must be sponsored by a PGC PI and analysts must sign the PGC analyst memo.
Yes. Any proposal involving more than one phenotype is considered cross-disorder research. Investigators pursuing cross-disorder research are encouraged to share their proposals with the CDG to avoid duplicate analyses and so that the proposals can be tracked and made available to PGC members interested in participating in ongoing analyses.
Once you develop a secondary analysis proposal, it should be submitted to the chair(s) of whichever working group(s) the proposed analysis includes. Each working group has its own procedures for vetting of secondary analysis proposals; please contact the workgroup chair for more information.
Proposals should be submitted to jsmoller@mgh.harvard.edu through a CDG representative of one of the workgroups whose data will be involved. All PGC members may propose cross-phenotype analyses. Approval by the participating Workgroups is required. The CDG maintains a list of these analyses and their primary contacts on a Google doc so that PGC members interested in participating may contact the proposing investigators. To avoid duplicative analyses, investigators pursuing cross-phenotype analyses are encouraged (but not required) to share their proposals with the CDG so that they can be tracked and to enter their proposal on this shared site. Submissions are typically discussed on the next scheduled CDG Workgroup call and we ask that a representative of the proposal join the call to briefly (10-15 minutes max) present the proposal.  In addition, secondary analyses are presented periodically on CDG Workgroup calls to track their progress and provide input from CDG members and other attendees. The CDG teleconferences, open to all PGC members, are a forum and resource for scientific discussion and consultation regarding cross-phenotype analyses. 
CDG conference calls are typically held on the first Wednesday of the month. It is best to submit your CDG proposal before these calls for discussion.
This varies by working group and depends on the frequency with which the workgroup has teleconferences. Simple proposals could be approved within weeks while proposals that require greater discussion may have to wait until the group meets via teleconference. 

For cross-disorder analysis, approval by the disorder workgroups contributing data is required but approval by the CDG is not strictly required.  Cross-disorder proposals are typically discussed on the next scheduled CDG call after submission. If the next CDG is not anticipated within one month, the proposal will be circulated to the CDG by email for any comments and can be considered approved if no concerns are raised within 1 week of email distribution.  

FAQ - About Data Access

ANY team members who will be conducting analysis (i.e., have their hands on data) need their own login credentials. You may not share your login details with anyone else, including members of your team. Sharing login credentials creates points of vulnerability opening LISA up for a possible security breach. There are ways to tell if credentials have been shared and the LISA administrators do regular checks for this. PLEASE DO NOT SHARE YOUR LOGIN INFORMATION UNDER ANY CIRCUMSTANCES.
dbGaP approval can only be obtained by independent principal investigators with an eRA commons ID. dbGaP approval is both investigator and institution specific. This means that PIs at the same University or institution who are collaborating on a project, may apply together for dbGaP access. PIs at different institutions who are collaborating on a project must have their own dbGaP access approval through their institution. Students, postdocs, and research staff are approved through their supervisor’s application and do not need to be named on the proposal. Please note that the PI takes personal responsibility for the proper use of dbGaP data by ANY lab member.
After the secondary analysis proposal has been approved, you can apply for access to individual level data. Please visit the Data Access page and click on “How to Request Individual Level Data.”
To apply for dbGaP access you must have an eRA account, an institutional signing official, and a short project proposal. The PGC has created a sample tutorial on applying to dbGaP that can be found here under the dbGaP Access heading. More information can be found on the dbGAP webpage under “Controlled Access Data.”
Instructions on how to apply for access to data through the NIMH Data Repository can be found here.
Consult the PGC DAC representative for the workgroup that you wish to work with. The DAC representative will let you know who to contact and what information is needed for securing permission to special permission data sets.
You must apply for a LISA account through the Surfsara website. This typically takes one week to ten days.
For each workgroup, there is a “fast-track” dataset which may be obtained upon approval of the secondary analysis proposal. Additionally, there are datasets available for request after obtaining permission from outside entities (i.e., dbGAP, NIMH, individual investigators). If you are unfamiliar with the dataset requirements, it is a good idea to contact your DAC representative before submitting your secondary proposal. He or she will let you know what permissions are required. Once you have obtained permissions, the Data Access Portal will allow you to submit these materials for review. Please be aware that all uploaded documentation must be in PDF or Word format.
You will select the datasets that you wish to access in the Data Access Portal. After your submission is approved by the DAC, an automated email is sent to the DAC co-chairs and the LISA helpdesk instructing the LISA administrators to add you to the unix groups that match the datasets you requested. Once this is done, you will receive an email alerting you to the presence of a new README in your home directory. The README contains instructions on how to create symbolic links to the data you have requested.
The goal of the PGC DAC is to facilitate access to data while ensuring we are all in compliance with privacy mandates. At any point if you are lost or stuck, you may contact your workgroup DAC representative, the , your workgroup chair, or the .
Please see the “How To …” tab on the Data Access page.
As the lab PI, you are responsible for the security of the data and must obtain permission for your analyst to access the data. This means YOU must: 1) submit the secondary analysis proposal and name your analyst on the proposal 2) request any additional permissions (ie., dbGAP, NIMH, individual investigators) for data sets your lab will request 3) sign the PGC memo of understanding. Your ANALYST must: 1) sign the PGC analyst memo 2) sign the WTCCC data use agreement 3) obtain a user account on LISA 4) submit the request through the Data Access Portal and provide evidence of PI permissions