Recalibrating ISTDs using NIST

Errors - in calculations and quantifications

SLING Research Updates 2023

Bo Burla - Singapore Lipidomics Incubator - National University of Singapore

Calculation Error

OH NO - wrong by a factor 2 ?!

→ Plasma and ISTD volume ratio was not considered in the calculation…

Unpublished data

Let’s fix the error - and (again) OH NO !?

→ Well, where to apply this factor 2 into the calculation (= R script) ?

Unpublished data

Calculation Errors at Data Processing

  • 34 labs participated and read the manuscipt
  • 1 lab highlighted the error
  • Would reviewers have realized it?


  • Analysts bear the responsibility
  • Think about risks of data processing
  • Include and analyse test datasets with manually curated results


  • Communucation between lab analysts and data analysts

Quantitation Error

Quantitation using Cer d18:1/16:0-[D31]


Analysis Dec 2017
n = 22 (SG)
Lipid Species Plasma Concentration
Cer d18:1/16:0 Study Samples 0.29 µmol/L
NIST SRM1950 0.41 µmol/L
NIST SRM1950
Cer d18:1/24:0 Study Samples 4.37 µmol/L
NIST SRM1950 3.70 µmol/L
NIST SRM1950
Analysis April 2019
n = 28 (CH)
Lipid Species Plasma Concentration
Cer d18:1/16:0 Study Samples 7.94 µmol/L
NIST SRM1950 7.60 µmol/L
NIST SRM1950
Cer d18:1/24:0 Study Samples 91.0 µmol/L
NIST SRM1950 75.0 µmol/L
NIST SRM1950

Quantitation using Cer d18:1/16:0-[D31]


Analysis Dec 2017
n = 22 (SG)
Lipid Species Plasma Concentration
Cer d18:1/16:0 Study Samples 0.29 µmol/L
NIST SRM1950 0.41 µmol/L
NIST SRM1950 (Ref) 0.24 µmol/L
Cer d18:1/24:0 Study Samples 4.37 µmol/L
NIST SRM1950 3.70 µmol/L
NIST SRM1950 (Ref) 2.41 µmol/L
Analysis April 2019
n = 28 (CH)
Lipid Species Plasma Concentration
Cer d18:1/16:0 Study Samples 7.94 µmol/L
NIST SRM1950 7.60 µmol/L
NIST SRM1950 0.24 µmol/L
Cer d18:1/24:0 Study Samples 91.0 µmol/L
NIST SRM1950 75.0 µmol/L
NIST SRM1950 2.41 µmol/L
  • Ceramide d18:1/16:0-[D31] degraded to 3% of initial response within 1.5 years !
  • Concentrations of all species normalized with this ISTD will be off (by the same factor)

Using RM to re-calibrate ISTDs

Triebl et al. (2020)

  • Divide concentration in study sample by concentration in NIST SRM1950
  • Unit: relative concentrations (relative to NIST), except when abs, concentration in known

ISTD recalibration using RM (e.g. SRM 1950 or LTR)

  • Calculate ratio between measured and expected concentration of (idially authentic) species in a RM (e.g. NIST)
  • Apply this ratio (correction factor) at quantitation for all species normalized with this ISTD
  • Unit: ‘absolute’ concentration, based on the NIST reference dataset
  • Can/could be used also for ISTDs used to normalize lipids of another class

Another study with ceramides

  • Using deuterated ISTDs, still needed NIST re-calibration
  • Requirement: NISTs to be co-processed exactly as the study samples!

Quantitation using re-calibrated ISTD


Analysis Dec 2017
n = 22 (SG)
Lipid Species Plasma Concentration
Cer d18:1/16:0 Study Samples 0.25 µmol/L
NIST SRM1950 0.24 µmol/L
NIST SRM1950 (Ref) 0.24 µmol/L
Cer d18:1/24:0 Study Samples 2.48 µmol/L
NIST SRM1950 2.41 µmol/L
NIST SRM1950 (Ref) 2.41 µmol/L
Analysis April 2019
n = 28 (CH)
Lipid Species Plasma Concentration
Cer d18:1/16:0 Study Samples 0.17 µmol/L
NIST SRM1950 0.24 µmol/L
NIST SRM1950 0.24 µmol/L
Cer d18:1/24:0 Study Samples 2.17 µmol/L
NIST SRM1950 2.41 µmol/L
NIST SRM1950 2.41 µmol/L

References

Triebl, A., Burla, B., Selvalatchmanan, J., Oh, J., Tan, S.H., Chan, M.Y., … Wenk, M.R. (2020). Shared reference materials harmonize lipidomics across MS-based detection platforms and laboratories. Journal of Lipid Research, 61(1), 105–115. DOI: 10.1194/jlr.D119000393