Application of Deep Learning in Biological Data Compression

TL;DR

This paper explores a novel deep learning-based method using implicit neural representations to efficiently compress Cryo-EM biological data, balancing storage size and reconstruction quality for research and educational use.

Contribution

It introduces a new approach combining neural networks, positional encoding, and weighted loss to improve Cryo-EM data compression over traditional methods.

Findings

Effective compression of Cryo-EM data achieved

Enhanced reconstruction accuracy with positional encoding

Balanced compression ratio and data fidelity

Abstract

Cryogenic electron microscopy (Cryo-EM) has become an essential tool for capturing high-resolution biological structures. Despite its advantage in visualizations, the large storage size of Cryo-EM data file poses significant challenges for researchers and educators. This paper investigates the application of deep learning, specifically implicit neural representation (INR), to compress Cryo-EM biological data. The proposed approach first extracts the binary map of each file according to the density threshold. The density map is highly repetitive, ehich can be effectively compressed by GZIP. The neural network then trains to encode spatial density information, allowing the storage of network parameters and learnable latent vectors. To improve reconstruction accuracy, I further incorporate the positional encoding to enhance spatial representation and a weighted Mean Squared Error (MSE) loss function to balance density distribution variations. Using this approach, my aim is to provide a practical and efficient biological data compression solution that can be used for educational and research purpose, while maintaining a reasonable compression ratio and reconstruction quality from file to file.