Representability of genomic signals

Area of Science:

• Bioinformatics
• Data Visualization
• Signal Processing

Background:

• Traditional data representation methods face challenges with complex genomic signals.
• Visualizing large-scale genomic data requires effective and scalable techniques.

Purpose of the Study:

• To introduce a novel approach for data representability using relative data variation.
• To analyze the representability of typical signals, with a focus on genomic data.
• To determine appropriate visualization methods for genomic signals across different scales.

Main Methods:

• Developed a novel approach based on relative data variation for pixel representation.
• Analyzed typical signal cases, emphasizing genomic signals.
• Investigated the visualization of genomic data variation along nucleotide sequences.

Main Results:

• The proposed method effectively visualizes genomic data variation as simple lines at low and large scales.
• For medium scales (hundreds of samples per pixel), traditional statistical data representation methods are necessary.
• Identified the cumulated and unwrapped phase of complex genomic signals as key visualization features.

Conclusions:

• Relative data variation offers a simplified visualization for genomic signals at certain scales.
• A hybrid approach combining relative variation and statistical methods is optimal for comprehensive genomic data analysis.
• The study provides insights into scalable and effective genomic data representation and visualization.