Cosmic Space Proteins

Hidden Proteins, Cosmic Rays, and AI's Growing Role in Scientific Discovery

An international research consortium identified 1,700 previously unknown proteins in what scientists had long dismissed as the human genome's 'dark' regions — stretches of DNA not expected to encode functional proteins. The microproteins challenge foundational assumptions about genetic organization and, researchers suggest, could yield new cancer therapy targets.

Separately, Penn State researchers studying ultrahigh-energy cosmic rays found that the most powerful particles reaching Earth may carry atomic nuclei heavier than iron. Their simulations indicate that ultraheavy nuclei lose energy more slowly across cosmic distances, potentially explaining how these particles maintain energy levels impossible to recreate in terrestrial experiments — offering a window into physical processes that remain poorly understood.

Artificial intelligence is accelerating both fields. New AI frameworks can extract precise distance and stellar property data from single supernova images alone, replacing observational campaigns that previously required multiple data types and significant telescope time. Researchers caution, however, that systems trained on existing data may miss phenomena that fall outside their training patterns — a limitation that applies as much to biology as to astrophysics.

The protein discoveries carry a broader epistemological implication: if 1,700 functional proteins were hiding in regions long labeled as genomic junk, other foundational assumptions about biological systems may similarly be awaiting revision. That uncertainty has direct consequences for drug development and genetic medicine, fields that have invested heavily in models built on the genetic maps now shown to be incomplete.