Hacks_E2_predictor

This code demonstrates 17b-estradiol (E2) prediction based on Hack LM et al. (2021) using a public dataset (GSE176394).

Hack LM, Nishitani S, Knight AK, Kilaru V, Maddox SA, Seligowski AV, Jovanovic T, Ressler KJ, Smith AK*, Michopoulos V, Epigenetic prediction of 17β-estradiol and relationship to trauma-related outcomes in women, Compr Psychoneuroendocrinol, 2021, 9;6:100045. PMID: 35757356 PMCID: PMC9216622 DOI: 10.1016/j.cpnec.2021.100045

Instructions

1. Download the dataset "GSE176394_Novakovic_GAHT_beta_values.csv.gz" from the following site. https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE176394
2. Unzip the .gz file with the appropriate software.
3. Download the other files, "Hack's E2 predictor.R", "FINAL_RF_70.30Split.rds", and "Hack_35probes.csv", uploaded here.
4. Put these three files and the unzipped dataset "GSE176394_Novakovic_GAHT_beta_values.csv" in the same directory.
5. Open the R code "Hack's E2 predictor.R" and run.
6. R packages "data.table", "wateRmelon", and "randomForest" will be required.