A compression model for protein structure comparison
Jerry Xu · Lexington, Massachusetts
Fifth Place ($90,000) · Regeneron Science Talent Search · 2026
A model that compresses a protein's molecular features into a compact numerical representation, so that structures can be compared efficiently without losing the features that matter.
Compression is only interesting if you can say what survives it. The claim here is not smaller — it is smaller while preserving the properties comparison depends on, and that is a testable statement.
- Protein feature extraction
- Learned compression to numerical strings
- Structure comparison benchmarking
- Information-loss evaluation
- For any compression or embedding, state what is preserved and demonstrate it — size alone is not a result
- Representation learning projects need a downstream task to be evaluated on
Archive entries are summaries compiled for study. Read them for how a project was structured, argued, and defended — not as a template to reproduce. Copying someone else's project is the one thing that will end yours.
Compiled by ResearchForge from the Society for Science public award announcement linked on this record. Names, hometowns, project titles and reported results are as published. The "why it worked" and "what to take from this" notes are ResearchForge's editorial reading of that public record — they are not statements by the students and not the judges' rationale. All rights to the original projects remain with their authors; no project materials are rehosted here.