ImmClass2019

ImmClass2019 classifies ten immune cell types and five T helper cell subsets and derives discriminative gene signatures from bulk RNA-seq and single-cell RNA-seq (scRNA-seq) data using elastic-net logistic regression.

Key Features:

• Elastic-Net Logistic Regression: Employs elastic-net logistic regression integrating L1 and L2 regularization to select gene subsets and estimate coefficients.
• High-dimensional Data Handling: Addresses challenges of high-dimensional transcriptomics data including low sample sizes, noise, and missing values.
• Comprehensive Classification: Constructs classifiers for ten immune cell types and five T helper cell subsets from RNA-seq data.
• Validation with scRNA-seq: Validates classifiers and gene signatures using single-cell RNA-seq (scRNA-seq) datasets to corroborate bulk RNA-seq results.
• Benchmarking: Benchmarks derived gene signatures against existing signatures to assess comparative performance.

Scientific Applications:

• Disease Mechanism Studies: Predicts the extent and functional orientation of immune responses to inform mechanistic studies in diseases such as cancer.
• Therapeutic Response Analysis: Assesses immune cell dynamics to analyze responses to therapeutic interventions.
• Transcriptomic Profiling: Provides classifiers and gene signatures for interpretation of bulk RNA-seq and scRNA-seq transcriptomic data.

Methodology:

Applies elastic-net logistic regression with L1 and L2 regularization for gene subset selection and coefficient estimation from RNA-seq data, validates models using scRNA-seq datasets, and benchmarks signatures against existing ones.