This folder contains code in R and C++ for replicating the analysis on Serological Signatures of SARS-CoV-2 infection as reported here. https://www.medrxiv.org/content/10.1101/2020.05.07.20093963v2
SARS-CoV-2_serology.csv A csv file containing measurements of antibody responses from our Luminex MAGPIX multiplex assay.
Fig1_single_AB.R An R script for preliminaryof antibody responses, plotting ROC curves, calculating AUCs, and calculating correlations.
Fig2_multi_AB.R Analysis of the assocation between multiple measurements of antibody response and SARS-CoV-2 infection status. A random forests classification algorithm (with cross-validation) is used.
Fig3_AB_kinetics.R R script taking the posterior outputs of the C++ code for antibody kinetics and plotting antibody levels over time and sensitivity.
Fig4_kinetic_classifier.R R script for model predicted sensitivity over time using a single antigen (spike) and multiplex combinations analysed with random forests algorithms. Depends on C++ output.
Fig5_sero_surveillance.R R script for statistical assessment of the trade-off between sensitivity and specificity on serological surveys.
SupFig1_crosspanel_validation.R R script for quantification of uncertainty of serological classification.
SARSCoV2_IgG_antigen_combination.xlsx Excel spreadsheet detailing classification performance of all combinations of IgG responses to 7 SARS-CoV-2 antigens using a random forests classifier.
Source.cpp C++ code for Bayesian statistical inference of anntibody kinetic models.
com.cpp linpack.cpp randlib.cpp randlib.cpp Randlib library for generation of random numbers for MCMC algorithm.
Chain_diagnostics.R R script for plotting the outputted MCMC chains from C++.
Stri_IPP_IgG_dil.txt RBD_IPP_IgG_dil.txt S1RBD_NA_IgG_dil.txt S1_NA_IgG_dil.txt S2_NA_IgG_dil.txt NP_IPP_IgG_dil.txt NP_NA_IgG_dil.txt Input files for C++ statistical inference. Note that these files were genereated by prcoessing SARS-CoV-2_serology.csv.