OHDSI Home | Forums | Wiki | Github

Who is working with UK Biobank?


(Vojtech Huser) #22

There was a meeting of the UK BB workgroup last week. It is now a subgroup of clinical study WG.

I volunteered to do all date variables in the prioritized list.

We completed some of them. There are two types: one that maps to a CDM column (date of birth) for person and one that maps to a EAV item (lab test).


(Michael Cantor) #23

Great to see this work progressing. The UKB pharma consoritum is also working on converting the UKB clinical data to OMOP, and are in the process of finalizing our agreement with a target start date in May. We expect a 4 month timeline and will make the mappings freely available to the scientific community.


(Maxim Moinat) #24

Great, could you share this already, either on the shared Drive or Github? Btw, we are not yet officially part of the Clinical Trial WG. Once that is sorted, we will formalise this working group with a wiki page containing all necessary info.

Would be great if they can join our effort, to avoid work duplication. Could you get us in touch with that group?


(Gregory Klebanov) #25

Hi Maxim - we are involved in helping consortium with BioBank. Let’s chat?


(Spiros Denaxas) #26

Hey Gregory, could you join our next UKB working group meeting? that would be great


(Spiros Denaxas) #27

Is it worth starting a wiki page now with our names and contact details?


(Maxim Moinat) #28

Thanks! I will send you the invite for our next call.

Yes, I do think so. We can start off with a separate wiki page and merge in the future with e.g. the CT working group.


(Maxim Moinat) #29

Tomorrow at 16h CEST we will have our third working group call. If you are willing to join, but are not invited, please let me know. @mcantor2 @SCYou @gregk


(Michael Cantor) #30

Thanks Maxim- I would like to join. Can you post the dialin information here?


(Dave Barman) #31

Hi Maxim, please add me as well. dave.barman@odysseusinc.com


(Andrew Williams) #32

Me too please.


(Maxim Moinat) #33

Of course, please use this link to dial in (16h CEST): https://meet.google.com/emq-pwsn-uhe


(Vojtech Huser) #34

The group just met. I was asked to put a proposal/post/discussion (about next steps)

  • Execute Achilles on currently converted data (hypothesis: is Achilles useful to data recipients)

  • in addition to work-by-data-element is to work-by-OMOP-table - example https://github.com/thehyve/ukbiobank-omop-etl/blob/master/docs/visit_occurrence.md

  • 2 workstreams: EHR data conversion and research data conversion

  • (just VH’s proposal) : in 4 weeks, extend the scope of the group to all registries (as informatician I must always strive to generalize most projects I work on) (rename UK biobank group to registries group and ukbb becomes just case study (we do AllOfUs next or Catalan SIDIAP: https://www.sidiap.org/index.php/database/general-details)

  • next meeting must be on platform that supports recording!


(Maxim Moinat) #35

Thanks Vojtech.

For that we first need a clear idea of who has already some UKB data mapped. I created a quick overview. @mcantor2 @kyriakosschwarz @Daniel_Prieto @Vojtech_Huser, please check whether this is correct and if you can extend this.

Funder Lead Supported by Primary Focus Started conversion? Current Mapping status
BigData@Heart Spiros Denaxas UCL/TheHyve Research+EHR no -
Regeneron Michael Cantor IQVIA/Odysseus EHR? no -
NIH/NLM Vojtech Huser ? no -
University of Zurich Kyriakos Schwarz ? ? ?
University of Oxford Dani/Albert ? ? ?

Agreed, I’ll think about how to make this into concrete actions for the next call.

I like the idea, but let’s not loose focus on UKB too soon. I do see that we can use the experience with UKB to set some more general mapping conventions to be used across UKB, AllOfUs and SIDIAP.


(Daniel Prieto-Alhambra) #36

So we (me and @Albert_Prats) have a couple data cuts, but we are not data custodians. we have access to UKB + linked datasets, but have mapped nothing. We do of course have some experience mapping other UK datasets (eg CPRD) that should be somewhat similar to the linked EMR, but have not done any mapping of UKB


(Oleg Zhuk) #37

Hi, @MaximMoinat, can you please add me to the next WG call as well?

Do you have any notes taken besides this theme? Or any other way of tracking progress?


(Alexander Davydov) #38

During the recent UK Biobank WG meeting we discussed the plan for implementation of UKB Survey data in the OMOP Vocabulary. The non-survey part looks less challenging so the same logic can be used in there.

Here is the detailed presentation and key points that affect the general OMOP vocabularies:

  • Answers/meas values to be standardized among the vocabularies and new relationship_id ‘Has Standard answer’ to be created and used for Standardized answer lists. Also requested here.
  • De-standardize and split ‘History of…’ concepts. The discussion and a whole perspective list of things to be split is here.
  • New ‘History of…’ concepts enriched with time context. Discussion/proposal here.
  • New Mapping table. Discussion is here. But it looks like we’re really good with additional synthetic pre-coordinated concepts.

@MaximMoinat @Christian_Reich @Vojtech_Huser @Andrew @mcantor2 @QI_omop @Dymshyts @Dave.Barman @Alexandra_Orlova @zhuk


t