# SDAMS

## Introduction

SDAMS is designed for differential abundance analysis for metabolomics and proteomics data from mass spectrometry. These data may contain
a large fraction of zero values and non-zero part may not be normally
distributed. SDAMS considers a two-part semi-parametric model, a logistic
regression for the zero proportion and a semi-parametric log-linear model for
the non-zero values. A kernel-smoothed likelihood method is proposed to estimate
regression coefficients in the two-part model and a likelihood ratio test is
constructed for differential abundant analysis. This package can be downloaded at Bioconductor.

## Sample R Code

R code and data files to perform all the analysis in our manuscript can be downloaded here.

- 'Real Data' folder contains prostate cancer proteomics data and lung cancer exosomal lipids data;
- 'RData' foler contains all simulation results used in both manuscript and supplementary materials;
- 'Rcode' foler contains all R functions to generate figures and tables in the paper.

Note:
- Run simulations using main simulation functions "control_mimic1", "control_mimic_DL1" and "control_real1";
- Generate figures using main plot functions "control_mimic_plot1" "control_mimic_plot2" and "control_real_plot1";
- Figure 2 in main paper was generated by function "plot_TPR";
- Table 1 in main paper was generated by function "plot_ROC";
- Figure 3 in main paper was generated by function "plot_FDR";
- Figure 4 and 6 in main paper were generated by function "plot_bar";
- Figure 5 in main paper was generated by function "plot_venn";
- Figure S1, S2, S7, S8 and S9 in supplementary materials were generated by function "plot_TPR";
- Figure S3, S4, S10, S11, and S12 in supplementary materials were generated by function "plot_FDR";
- Figure S5, S6, S13, S14 and S15 in supplementary materials were generated by function "plot_bar";
- Table S2 in supplementary materials was generated by function "control_type1";