Low cost, high performance processing of single particle cryo-electron microscopy data in the cloud.

Area of Science:

• Structural Biology
• Biophysics
• Computational Biology

Background:

• Advancements in electron microscopy enable high-resolution structure determination.
• High-performance computing is essential for cryo-electron microscopy (cryo-EM) data processing.
• Access to computing clusters can be a bottleneck for cryo-EM researchers.

Purpose of the Study:

• To develop an accessible, cloud-based computing environment for single particle cryo-EM.
• To reduce computational barriers for cryo-EM structure determination.
• To facilitate wider adoption of cryo-EM in structural biology.

Main Methods:

• Developed a cloud computing environment using Amazon's Elastic Compute Cloud (EC2).
• Integrated essential single particle cryo-EM software packages.
• Enabled the creation of scalable computing clusters (16-480+ CPUs).
• Validated the environment using a publicly available yeast ribosome dataset.

Main Results:

• The cloud environment provides a cost-effective solution for cryo-EM structure determination, estimated at $50-$1500 per structure.
• Processing times are comparable to local computing clusters.
• Demonstrated the feasibility of determining high-resolution structures using the cloud platform.

Conclusions:

• Amazon's cloud computing infrastructure offers a viable and accessible platform for single particle cryo-EM.
• This approach can democratize access to high-resolution structural biology tools.
• Facilitates broader research in structural biology by overcoming computational limitations.