/nicotine_metabolomics

Primary LanguageR

NMR Assay Dataframe Curation

Crystal R. Musser, M.S. (Wet Lab Manager, Katrina Claw Lab @ CU Anschutz, Department of Biomedical Informatics)

v.05/16/2024

PURPOSE:

Combine quantified LC/MS nicotine metabolite data with phase survey self-reported data, from specified cohort.

OBJECTS:

"t_allSamplesMetabolites" for curated nicotine metabolite dataset "t_allSamples" as the master (all variables all samples dataframe, post-QC, after specific exclusions) "clean_df" as the finalized, shared dataset, 18 variables, n=816 (inc. 28 controls)

SUPPLEMENTARY FILES to this script:

Data dictionaries (Supplementary File 1) "Clean" variable consensus/logic UML (Supplementary File 2)

REQUIRED INPUTS: 4

  1. combined experimental LC-MS/MS calculated and QC'd data in [ng/mL] serum concentrations for: , Cotinine, 3-hyroxy-Cotinine, Nicotine, CO, NO, Nornicotine (in this order) ---> "v6RawData_combined.csv" 2, 3) cohort survey results with associated data dictionary for interpretation ---> "phase2.xlsx" and "smoking.xlsx"
  2. phase I survey results with associated data dictionary for interpretating age, gender, smoking histotry, height, weight ---> "phase1.xlsx"

OUTPUT FILES GENERATED:

  1. A list of all combined results for all 824 samples; t_allSamples_postExclusions, file = "SHSph2_fullDF_05152024CRM.csv"
  2. A "clean" dataset with 18 variables for the final 816 participants to use for EWAS/GWAS; clean_df, file = "SHSph2_NMRwSurveyDF_clean05152024CRM.csv"
  3. A list of just sampleIDs of final 816 participants; clean_df$idNo, file = "SHSph2_IDs_clean05152024CRM.csv"

NOTE:

  • I'm bringing in calculated metabolic data using Kaja's data output from Teddy's script referencing Laura's library, and am assuming reported values pass QC
  • I intend on chaning the hardcoded bits into variables for scalability in a larger bioinformatic pipeline (dev in progress, 05/15/2024 CRM)

This script:

Calculated metabolite concentration data > Dataframe curation Script > Data Cleaning Script

Connecting it from:

Raw experimental LC-MS/MS data > metabolite standard curve generation > concentation back-calculations

Connecting it to:

Data Analysis Script > Figures & Tables Script

STILL TO DO:

  1. write in a check that there's not data being overwritten for that sample's concentration variable (ie reruns in the same file), handle this how?
  2. join with samples on my rerun list .xlsx