search

micronutrientsupport / database-architecture

The Postgres database code for the MAPS tool
3 stars 0 forks source link

List of biomarker attributes #4

Closed rbroth closed 1 year ago

rbroth commented 3 years ago

What nutrients do we get from biomarker data? Zinc? iondine?

fannymawbey commented 3 years ago

Most commonly:

Blood:

Hemoglobin Ferritin Soluble Transferrin Receptors (sTfR) Retinol Binding Protein (RBP) Serum Retinol Folate Vitamin B12 Zinc C-reactive Protein (CRP) Alpha 1 acid GlycoProtein (AGP)

Urine Iodine

rbroth commented 3 years ago

What are the units for those?

fannymawbey commented 3 years ago

I'll put everything in an excel file with upper and lower values when available.

rbroth commented 3 years ago

To how many significant figures and places behind the decimal point are folate, vitB12, and iodine? I've taken a stab at the others based on the example data in sample raw data Malawi WRA.xlsx but it might be worth you having a look over them. 

    , haemoglobin             numeric(5,2)
    , ferritin                numeric(5,2)
    , stfr                    numeric(3,2)
    , rbp                     numeric(3,2)
    , serum_retinol           numeric(3,2)
    , folate                    numeric(10,5)  -- ????
    , vitamin_b12               numeric(10,5)  -- ????
    , zinc                    numeric(3,1)
    , crp                     numeric(3,2)
    , agp                     numeric(3,2)
    , iodine                    numeric(10,5)  -- ????

The first number indicates how many significant digits a measurement has in total, while the second one indicates how many of those digits are to the right of the decimal point, e.g. Sensible measurements for haemoglobin are e.g. 132.75 - something like 132.678496 would not make sense, because the measurement instruments are not that precise.

fannymawbey commented 3 years ago

Sorry I forgot to respond folate: 5,2 vit B12: 5,1 iodine: 5,2

Just a question. will that block us to enter higher or more precise values if we want to? Or is that just an indication?

Thanks Fanny

rbroth commented 3 years ago

It'll throw an error and block data entry, the idea being that if someone (or rather, a computer program somewhere) is trying to enter iodine data with 8 figures behind the decimal point, the data is likely to be nonsense

rbroth commented 3 years ago

I assume that both rbc folate and serum folate have 3 digits before the decimal pont and 2 digits after it?

fannymawbey commented 3 years ago

RBC folate 3 before, 1 after

serum folate 2 before, 1 after

is that ok?

Le lun. 19 avr. 2021 à 15:27, Roman @.***> a écrit :

I assume that both rbc folate and serum folate have 3 digits before the decimal pont and 2 digits after it?

— You are receiving this because you were assigned. Reply to this email directly, view it on GitHub https://github.com/micronutrientsupport/database-architecture/issues/4#issuecomment-822510187, or unsubscribe https://github.com/notifications/unsubscribe-auth/AO7ZVAUG63PVVAGKK6PHA2TTJQ4VXANCNFSM4SIVG5JA .

rbroth commented 3 years ago

Yep, that's fine

rbroth commented 3 years ago

@fannymawbey Can you tell me more about the serum/plasma distinction mentioned in https://github.com/micronutrientsupport/database-architecture/issues/142#issuecomment-829112244?

we have to be specific. Sometimes they will collect serum, sometimes plasma, and we need to use the name of the one used by the survey. I think the idea that we had, as shown in the 'threshold' file, is that you will have a micronutrient 'folate' and then a matrix 'plasma, serum, red blood cell, whole blood, urine, breastmilk'.

Do we only to report which one was used so that the end user knows it, or do the R scripts give different results depending on whether you input serum folate vs. plasma folate?

fannymawbey commented 3 years ago

It's just so that the users know. The thresholds are the same whether it's plasma or serum.

On Thu, 29 Apr 2021, 2:53 pm Roman, @.***> wrote:

@fannymawbey https://github.com/fannymawbey Can you tell me more about the serum/plasma distinction mentioned in #142 (comment) https://github.com/micronutrientsupport/database-architecture/issues/142#issuecomment-829112244 ?

we have to be specific. Sometimes they will collect serum, sometimes plasma, and we need to use the name of the one used by the survey. I think the idea that we had, as shown in the 'threshold' file, is that you will have a micronutrient 'folate' and then a matrix 'plasma, serum, red blood cell, whole blood, urine, breastmilk'.

Do we only to report which one was used so that the end user knows it, or do the R scripts give different results depending on whether you input serum folate vs. plasma folate?

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/micronutrientsupport/database-architecture/issues/4#issuecomment-829256614, or unsubscribe https://github.com/notifications/unsubscribe-auth/AO7ZVAUNNCKM7RZ5BUBEQKDTLFQE7ANCNFSM4SIVG5JA .

rbroth commented 3 years ago

So we can have either serum retinol or plasma retinol, but the threshold values are the same? And we can have either serum folate or plasma folate, but the thresholds are the same? It sounds to me like these might be different types of sampled matter, so that we would have

Or am I thinking in the wrong direction here?

fannymawbey commented 3 years ago

I would do it like that as well. Note that for Red Blood Cell folate you would choose blood-whole for the matrix.

Le mar. 25 mai 2021 à 10:04, Roman @.***> a écrit :

So we can have either serum retinol or plasma retinol, but the threshold values are the same? And we can have either serum folate or plasma folate, but the thresholds are the same? It sounds to me like these might be different types of sampled matter, so that we would have

  • blood - serum
  • blood - plasma
  • blood - whole
  • urine
  • breastmilk

Or am I thinking in the wrong direction here?

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/micronutrientsupport/database-architecture/issues/4#issuecomment-847691249, or unsubscribe https://github.com/notifications/unsubscribe-auth/AO7ZVASEMZ35S54XXEVCETDTPNRYHANCNFSM4SIVG5JA .

rbroth commented 3 years ago

So to confirm: RBC-folate and Serum-folate are treated identically by the calculations. The only difference between them is that we want to tell the end-user that one is derived from whole-blood measurements and the other is derived from a "blood - serum" measurement?

rbroth commented 3 years ago

We're having a look at some biomarker data. It looks like the biomarker measurement data is already "pre-aggregated" at the measurement level. Instead of the data saying, "Here are the results for a blood sample from person x, and the results for a urine sample for person x", it's more like "here are biomarker measurements for person x. They were derived from urine and whole blood samples".

Does that sound right to you @fannymawbey ?

fannymawbey commented 3 years ago

sounds right, yes

Le ven. 28 mai 2021 à 11:29, Roman @.***> a écrit :

It looks like the biomarker measurement data is already "pre-aggregated" at the measurement level. Instead of the data saying, "Here are the results for a blood sample from person x, and the results for a urine sample for person x", it's more like "here are biomarker measurements for person x. They were derived from urine and whole blood samples"

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/micronutrientsupport/database-architecture/issues/4#issuecomment-850321251, or unsubscribe https://github.com/notifications/unsubscribe-auth/AO7ZVAVWIKZLQC3U45ZGO6DTP5WBFANCNFSM4SIVG5JA .

fannymawbey commented 3 years ago

Hemoglobin - whole blood (don't need to specify) Ferritin (serum or plasma) Soluble Transferrin Receptors (sTfR) (serum or plasma) Retinol Binding Protein (RBP) (serum or plasma) Retinol (serum or plasma) Folate (serum or plasma or red blood cell) Vitamin B12 (serum or plasma) Zinc (serum or plasma) selenium (serum or plasma) C-reactive Protein (CRP) (serum or plasama) Alpha 1 acid GlycoProtein (AGP) (serum or plasma)

Urine Iodine

rbroth commented 3 years ago

Does folate have different thresholds depending on whether it is red blood cell vs plasma vs serum? I'm noticing different numbers in the thresholds table

fannymawbey commented 3 years ago

Red blood cell folate VS plasma or serum will be different in terms of threshold. That's why it's so important to be precise.

Le mer. 2 juin 2021 à 10:37, Roman @.***> a écrit :

Does folate have different thresholds depending on whether it is red blood cell vs plasma vs serum? I'm noticing different numbers in the thresholds table

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/micronutrientsupport/database-architecture/issues/4#issuecomment-852873468, or unsubscribe https://github.com/notifications/unsubscribe-auth/AO7ZVAWMOVVBPI7NYAUO5XDTQX3ULANCNFSM4SIVG5JA .

rbroth commented 3 years ago

Do any of the other biomarkers have different thresholds depending on plasma/serum/wholeblood?

fannymawbey commented 3 years ago

No.

Le mer. 2 juin 2021 à 10:43, Roman @.***> a écrit :

Do any of the other biomarkers have different thresholds depending on plasma/serum/wholeblood?

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/micronutrientsupport/database-architecture/issues/4#issuecomment-852878574, or unsubscribe https://github.com/notifications/unsubscribe-auth/AO7ZVAXX6RGBV73DRCYYVLLTQX4KPANCNFSM4SIVG5JA .

bgsandan commented 1 year ago

Closing due to inactivity and probably completeness