best fields for blood draw time stamps for proteomics and metabolomics?
I have a question regarding the proteomic and metabolomic data in the UKB RAP.
Are there data fields on the Showcase that indicate when blood draws were performed for proteomics and metabolomics data? I?m guessing it could be https://biobank.ndph.ox.ac.uk/ukb/field.cgi?id=3166 or https://biobank.ndph.ox.ac.uk/ukb/field.cgi?id=21842, but I am just wondering if all of the data in metabolomics and proteomics instance 0 corresponds to blood draws performed in Instance 0 of "Time blood sample was collected" in Data Field 3166 or 21842?since there seem to be multiple instances? Furthermore, what is the difference between field 3166 and 21842? Is one more accurate than the other?
Thank you.
Comments
2 comments
Field 3166 has up to 7 different timestamps per participant for each instance, such as instance 0 (baseline visit), because the blood was collected in 7 separate containers. For more details on the blood collection protocol, see Resource 100227. There are even some pictures to illustrate this point in the other resources in Category 100002. Field 3166 should record the date and time that each container was scanned into the system. The 7 dates will always be the same, and only the times within the timestamps will be different. Any deviation from this will be an error of some kind.
Field 21842 has only one timestamp per participant per instance, and it is the date and time that the clinic staff member signed off the blood collection stage as completed. Thus, the time in 21842 should be shortly after the last of the times in 3166.
Field 53 has only a date (no time) per participant per instance, and it is the date that the participant attended the clinic.
If everything worked according to plan, the dates in fields 3166, 21842 and 53 should all match (for a particular participant and a particular instance). If there are any discrepancies, you might like to bear in mind that field 3166 relies on the scanning software being correctly set, and field 21842 relies on a certain amount of manual staff data input. There was a known issue with field 3166 having a clock error, and those timestamps affected have been removed. There may be small numbers of inconsistencies remaining.
All the Nightingale NMR metabolomics measurements were done using blood from either visit 0 or visit 1. If the assay value has an associated instance 0, this means that the blood used for the assay was taken from the participant during their visit 0, on the date that is in field 53_i0.
Approximately 1439 participants have metabolomics measurements for both their visit 0 and visit 1 samples. There are illustrations of the between-visit correlations in Resource 3003.
The ins_index value in the table olink_data (AMS-Showcase) indicates the visit. A few samples were assayed from instance 2 or 3 (imaging visits), most were from instance 0 (baseline). None are from instance 1. On the RAP, the olink_data table has been split into separate fields, the fields have an instance suffix, and the Entity incorporates the instance.
Within UKB, "instance" is most often (but not always) used to signify "visit". To confirm that it signifies "visit" for the metabolomics data, see eg field 23475 Acetate which says "Defined-instances run from 0 to 1, labelled using Instancing 2", and then see Instance 2. To confirm that it signifies "visit" for the proteomics data, see eg field 30903, similarly.
Thank you very much!
Please sign in to leave a comment.