MAI
A two-step approach to imputing missing data in metabolomics. Step 1 uses a random forest classifier to classify missing values as either Missing Completely at Random/Missing At Random (MCAR/MAR) or Missing Not At Random (MNAR). MCAR/MAR are combined because it is often difficult to distinguish these two missing types in metabolomics data. Step 2 imputes the missing values based on the classified missing mechanisms, using the appropriate imputation algorithms. Imputation algorithms tested and available for MCAR/MAR include Bayesian Principal Component Analysis (BPCA), Multiple Imputation No-Skip K-Nearest Neighbors (Multi_nsKNN), and Random Forest. Imputation algorithms tested and available for MNAR include nsKNN and a single imputation approach for imputation of metabolites where left-censoring is present.
- Repository
- github.com/kechrislab/mai
Source attribution
- Bioconductor — MAI
Related resources
Tools to harmonize bulk RNA-seq matrices, optionally apply batch correction, and train cross-validated classification models using ranger, glmnet, or xgboost. Supports leakage-safe feature selection, permutation importance, SHAP-based interpretability, and calibration methods (Platt or isotonic). Provides stability metrics across folds, embeddings (PCA/UMAP), ROC visualization, SHAP dependence plots, and tidy ranked-gene tables for downstream analysis.
MPRAnalyze provides statistical framework for the analysis of data generated by Massively Parallel Reporter Assays (MPRAs), used to directly measure enhancer activity. MPRAnalyze can be used for quantification of enhancer activity, classification of active enhancers and comparative analyses of enhancer activity between conditions. MPRAnalyze construct a nested pair of generalized linear models (GLMs) to relate the DNA and RNA observations, easily adjustable to various experimental designs and conditions, and provides a set of rigorous statistical testig schemes.
Non-parametric method for identifying differentially expressed (up- or down- regulated) genes based on the estimated percentage of false predictions (pfp). The method can combine data sets from different origins (meta-analysis) to increase the power of the identification.
The package offer different classifiers based on comparisons of pair of features (TSP), using various decision rules (e.g., majority wins principle).
This package provides a streamlined workflow for the quanTIseq method, developed to perform the quantification of the Tumor Immune contexture from RNA-seq data. The quantification is performed against the TIL10 signature (dissecting the contributions of ten immune cell types), carefully crafted from a collection of human RNA-seq samples. The TIL10 signature has been extensively validated using simulated, flow cytometry, and immunohistochemistry data.
An R package for integrated differential expression and differential network analysis based on omic data for cancer biomarker discovery. Both correlation and partial correlation can be used to generate differential network to aid the traditional differential expression analysis to identify changes between biomolecules on both their expression and pairwise association levels. A detailed description of the methodology has been published in Methods journal (PMID: 27592383). An interactive visualization feature allows for the exploration and selection of candidate biomarkers.