Eproteins

Eproteins are a term used in computational and theoretical biology to denote artificially evolved protein sequences generated by in silico evolutionary algorithms. They are designed to serve as model systems for exploring how sequence determines structure and function, without being constrained by the history of natural evolution.

Generation and properties: Eproteins are produced by iterative cycles of mutation, selection, and recombination in a computer model. Fitness criteria typically include predicted stability, folding tendency, and avoidance of aggregation, and sometimes functional proxies such as active-site geometry. The resulting sequences are often nonhomologous to known proteins. Predicted structures are generated by ab initio methods or structure prediction tools, enabling rapid assessment of foldability.

Applications: Researchers use Eproteins to benchmark protein design and structure prediction algorithms, to study principles of evolvability and robustness, and to test hypotheses about sequence-structure-function relationships under controlled conditions. They can help identify design features that promote soluble expression and reliable folding beyond what is observed in natural sequences.

Limitations: Eproteins are not known to exist in nature; findings may not translate directly to natural proteins due to model biases and missing cellular context. The approach relies on computational models and assumptions about fitness that may not capture all aspects of biophysics, binding, or regulation.

See also: protein engineering, protein design, computational biology, simulated evolution, structure prediction.