Training differential analysis results — TRAINING

Training summary statistics from differential analysis (DA) that tests the effect of training on each feature within each sex. One data frame per data type and tissue.

Format

A data frame with up to 47 variables:

assay: character, assay code used in data release. See MotrpacBicQC::assay_codes.
tissue: character, tissue code used in data release. See MotrpacBicQC::bic_animal_tissue_code.
feature_ID: character, MoTrPAC feature identifier
removed_samples: character, comma-separated list of vial labels excluded from differential analysis or NA
removed_samples_male: character, comma-separated list of male outliers (vial labels) removed from differential analysis
removed_samples_female: character, comma-separated list of female outliers (vial labels) removed from differential analysis
dataset: character, specific dataset
dataset_male: character, specific dataset
dataset_female: character, specific dataset
panel: character, LUMINEX panel for immunoassay analytes
full_model: character, full model used for differential analysis
full_model_male: character, full model used for differential analysis in males
full_model_female: character, full model used for differential analysis in females
reduced_model: character, effective reduced model used for differential analysis
reduced_model_male: character, effective reduced model used for differential analysis in males
reduced_model_female: character, effective reduced model used for differential analysis in females
covariates: character, comma-separated list of adjustment variables
covariates_male: character, comma-separated list of adjustment variables used for males
covariates_female: character, comma-separated list of adjustment variables used for females
fscore_male: double, F-statistic for males
fscore_female: double, F-statistic for females
lrt: double, likelihood ratio test statistic when only one sex exists
lrt_male: double, likelihood ratio test statistic for males
lrt_female: double, likelihood ratio test statistic for females
p_value_male: double, nominal p-value for males
p_value_female: character, nominal p-value for females
adj_p_value_male: double, IHW-adjusted p-value for males
adj_p_value_female: double, IHW-adjusted p-value for females
p_value: double, combined male and female nominal p-value using the sum of logs
adj_p_value: double, IHW-adjusted combined p-value; value provided by selection_fdr in the timewise summary statistics
tissue_abbreviation: character, tissue abbreviation, one of TISSUE_ABBREV
Chr: character, chromosome where feature is located. METHYL only.
Locus: character, name of feature. METHYL only.
EntrezID: integer, Entrez ID of corresponding gene. METHYL only.
Symbol: character, gene symbol of corresponding gene. METHYL only.
is_targeted: logical, whether metabolite is targeted. Metabolomics only.
site: character, Chemical Analysis Site (CAS) name. Metabolomics only.
metabolite: character, name of metabolite as appears in the CAS's data. Metabolomics only.
metabolite_refmet: character, RefMet name of metabolite. Metabolomics only.
groups_tested_male: character, timepoints used to perform the F-test in males. Some tissues or assays are missing timepoints. Metabolomics only.
groups_tested_female: character, timepoints used to perform the F-test in females. Some tissues or assays are missing timepoints. Metabolomics only.
mz: double, mass over charge. Metabolomics only.
rt: double, retention time. Metabolomics only.
neutral_mass: numeric, neutral mass. Metabolomics only.
cv: double, feature coffeicient of variation in the dataset. Metabolomics only.
meta_reg_het_p: double, heterogeneity p-value from meta-regression. A smaller p-value indicates more disagreement between platforms. Metabolomics meta-regression only.
original_ftest_ps: character, comma-separated list of original F-test p-values. Metabolomics meta-regression only.

Details

While the more commonly-used timewise summary statistics are available directly in this package (e.g., TRNSCRPT_DA, PROTEOME_DA, IMMUNO_DA, METAB_DA, METAB_DA_METAREG, TRAINING_REGULATED_FEATURES, REPFDR_INPUTS), training summary statistics are available for download from Google Cloud Storage. Use MotrpacRatTraining6mo::load_training_da() to download results for a given tissue and ome. Note that the adjusted p-value from the training differential analysis is already included in the timewise summary statistics as selection_fdr.

Training differential analysis was performed using either F-tests or likelihood ratio tests depending on the ome. Find the functions used to generate these results listed here. See details in the supplementary methods of the manuscript.