Multi-Modal Research Dataset Normalization Pipeline

Design a comprehensive data normalization strategy for heterogeneous scientific research datasets combining genomic sequencing, spectroscopic measurements, and environmental sensor readings. Create a robust preprocessing workflow that handles missing values, outlier detection, and cross-modal feature scaling using advanced statistical techniques. Develop a modular Python script that can automatically detect data types, apply appropriate transformations, and generate a standardized DataFrame with comprehensive metadata tracking.