covid mutation predictor

• 

  website mutation analysis page

Inspiration

We have observed the current and past conditions caused by COVID-19. This led us to consider that if we could predict mutations early, it would represent a major step forward in preventing the spread and impact of the virus.

What it does

Our website is designed to predict upcoming COVID-19 mutations and assess their risk score based on the protein sequence of the current variant.

How we built it

We utilized machine learning alongside API integrations, including:

NCBI E-utilities – Literature and sequence data (public access)

UniProt REST API – Protein sequences and annotations (public access)

PDB API – Protein structures (public access)

EMBL-EBI Proteins API – Protein variation data (public access)

Open-source tools – ColabFold, AlphaFold, NGL Viewer

The platform is designed to be modular and scalable, with each feature implemented as a focused, maintainable React component. It improves upon traditional medical supercomputing approaches by offering:

Real-time accessibility vs. batch processing

Interactive visualization vs. static reports

Collaborative features vs. isolated analysis

User-friendly interface vs. command-line tools

Integrated workflows vs. separate tools

Challenges we ran into

We encountered difficulties in achieving high prediction accuracy and resolving bugs during development.

Accomplishments that we're proud of

We successfully built a solid foundation that we can expand upon in the future.

What we learned

We gained insights into how COVID-19 mutations form and evolve.

What's next for covid mutation predictor

We plan to continue fixing bugs, improving accuracy, and enhancing the platform's overall performance.

Built With

• alphafold
• embl-ebi
• ncbi
• pdb
• uniprot