DROMA Components

Overview

DROMA_DB is the unified SQLite data resource in the DROMA ecosystem. It organizes heterogeneous drug response and multi-omics datasets into one harmonized database for efficient cross-project access.

Key Features

• Unified database structure for heterogeneous multi-omics resources
• Project-oriented architecture with 21 datasets
• Optimized SQLite with indexing for fast queries
• Comprehensive metadata management
• Support for cell line, PDC, PDO, PDX, and clinical-oriented resources

Database Content

Contains harmonized DROMA resources including:

# Create database
library(DROMA.Set)
createDROMADatabase(
  "droma.sqlite", 
  "data/"
)

Overview

DROMA_Set is the R data management layer of DROMA. It provides S4 classes for constructing, loading, and coordinating project-level and cross-project data objects from DROMA_DB.

Core Classes

Key Features

• Flexible data loading with filtering options
• Sample overlap detection across projects
• Comprehensive metadata management
• Database connectivity and efficient querying
• Support for all molecular profile types

# Install from GitHub
devtools::install_github(
  "mugpeng/DROMA_Set"
)

# Create DromaSet
gCSI <- createDromaSetFromDatabase(
  "gCSI", "droma.sqlite"
)

# Load data
gCSI <- loadMolecularProfiles(
  gCSI, "mRNA", c("BRCA1", "TP53")
)

Overview

DROMA_R is the main analysis engine in the DROMA ecosystem. It provides harmonized statistical analysis and visualization functions for drug response and omics association studies across datasets.

Analysis Methods

Core Functions

• AAC-based response handling with project-level z-score normalization
• Cross-project harmonized comparison across model systems
• Drug-omics association analysis across transcriptomic and other omics layers
• Batch biomarker screening and meta-analysis workflows
• Pathway and GSVA-style downstream interpretation support

Visualization Outputs

• Forest plots for meta-analysis results
• Volcano plots for batch analysis
• Scatter plots for continuous associations
• Box plots for discrete comparisons
• Publication-ready figure themes

Overview

DROMA_Meta is a workflow application package in the DROMA ecosystem. It does not replace DROMA_R; instead, it packages standardized multi-step biomarker discovery pipelines on top of DROMA.Set and DROMA.R.

What It Adds

• Workflow-level orchestration for standardized analysis runs
• Reusable multi-step biomarker discovery pipelines
• Script-level looping over drug and tumor-type combinations
• Structured output directories for reproducible runs
• Application-layer packaging rather than a new low-level API

# Install workflow layer
devtools::install_github(
  "mugpeng/DROMA_Meta"
)

Overview

DROMA_MCP is a Model Context Protocol server that bridges AI assistants with the DROMA ecosystem. It exposes dataset access, data loading, and analysis actions through natural language interfaces.

AI Capabilities

Integration Examples

# Install via pip
pip install droma-mcp

# Start server
droma-mcp run --db-path droma.sqlite

Overview

DROMA_Py provides a Pythonic interface to the DROMA ecosystem, enabling Python data scientists and bioinformaticians to access and analyze cancer pharmacogenomics data seamlessly. It bridges the gap between R-based DROMA components and Python workflows.

Core Capabilities

Python Integration

• Pandas DataFrame integration
• NumPy array support
• Type hints for better IDE support
• Comprehensive error handling
• Jupyter notebook compatibility

# Install from PyPI
pip install droma-py

# Basic usage
import droma_py as dp
db = dp.DROMADatabase("droma.sqlite")

# Load molecular data
data = db.load_molecular_profiles(
    "CCLE", "mRNA", features=["BRCA1"]
)

# Harmonize gene names
harmonized = dp.harmonize_gene_names(data)

Overview

DROMA_Web provides an intuitive, browser-based interface for drug response and omics analysis. Built with Shiny, it offers real-time interactive analysis without requiring programming knowledge.

Web Modules

User-Friendly Features

• No programming knowledge required
• Real-time interactive visualizations
• Dynamic filtering and data exploration
• Export results and figures
• Cross-platform browser compatibility