Family Data Examples from source (original) data submissions

[lopierra]

HTP:

• They provided their own Family Type enums and Proband designations
• For some reason the Proband ID is listed under "OtherFamilyMemberID" - this is not as intended; this col was meant for like grandparents, etc.
• F0084 and F0088 are examples of families with two Probands
• There are also larger control-only families, e.g. F0023 - these families don't have Probands

X01-Hakonarson:

• I'm not sure if there are any 2-proband families, but there are some more different relationships in this dataset
• e.g. KFDS0563717437 has grandparents, mother, and proband
• Note that many of the participants have 2 samples (DNA + RNA) so they will have 2 rows of data

DS-Connect IHQ (initial health questionnaire):

• They did not designate Probands (all of the participants in the study have DS)
• If there are 2 children enrolled in the same family, the Patient ID of the first one to enroll becomes the "Parent ID" for the 2nd sibling

DS-Sleep:

• They designated probands and family relationships
• There is one family with 2 probands (twins) - 113

TEAM-DS:

• All of their participants have DS
• They did not provide any proband or family information, but when asked, they informed me that there were 2 sibling pairs (101061/101062 and 101088/101089)

ABC-DS:

• They did not provide proband designations or family info, but they provided a separate mapping of participants (col B, ABCDS_ID_DS) and their matched sibling controls (col A, ABCDS_ID_SC). Most participants did not have sibling controls.

[lopierra]

Links from Matt:

• https://github.com/GA4GH-Pedigree-Standard/pedigree
• https://lucid.app/lucidchart/a96e614d-1099-4c3c-9902-afbc81049280/edit?viewport_loc=-4293%2C-1491%2C12378%2C6958%2C7-nKhd6nEzpu&invitationId=inv_95cb0794-fa79-4dce-9a2d-4bc1cd45f025

[lopierra]

We should also consider improving the enums for Family Type to be more descriptive, which would be especially helpful to summarize if we plan to provide a lot more detailed relationship data.

Currently we have:

• Control-only
• Proband-only
• Duo (1 parent + 1 child)
• Trio (2 parents + 1 child)
• Trio+ (2 parents + >1 children + other relatives)
• Other (Siblings only, 1 parent + >1 children, grandparents, etc.)

HTP has more descriptive Family Types for both T21 and D21 families - maybe something like what they're using?

• D21 Only
• D21-Multiple Sibs
• D21-Parent Duo
• D21-Parent-Multiple Sibs
• D21-Parents Trio
• D21-Parents-Multiple Sibs
• D21-Sib Duo
• T21 Only
• T21-Multiple Sibs
• T21-Other Trio
• T21-Parent Duo
• T21-Parent-Multiple Sibs
• T21-Parents Trio
• T21-Parent-Sib Trio
• T21-Parents-Multiple Sibs
• T21-Parents-Sib Quad
• T21-Sib Duo