Scientists at University of Wisconsin-Milwaukee in the US developed a new generation of powerful algorithms to reconstruct sequential images from an ocean of unsorted, noisy data.
Using the brightest and quickest imaging equipment available - an X-ray Free Electron Laser (XFEL) - researchers collected millions of individual "snapshots" of a virus in unknown orientations and states.
"In the past, scientists have tried to infer what's happening in a molecular-scale biological process by looking at a still photo at the start and a still photo at the end of a process," said Abbas Ourmazd, professor at UWM.
By combining concepts from machine learning, differential geometry, graph theory and diffraction physics, the researchers created an algorithm able to reconstruct sequential images.
More From This Section
In order to replicate, a virus invades a healthy cell, releases its DNA and effectively hijacks the cell's machinery to fabricate copies of itself.
The multitude of viral progeny is then expelled to infect other cells.
The study, published in the journal Nature Methods, shows that the virus re-arranging its genomic content and forming a tubular structure to empty its DNA into a cell.
In addition to showing how these events unfold, researchers discovered that the reorganisation of the virus' genome and the formation of a tubular structure are not independent events, but part of a concerted simultaneous process.
Most viruses are too small to be photographed by light. The XFEL's intense X-ray flashes produce "snapshots" of particles at the nanoscale through diffraction.
The X-rays hit the particle and scatter in a pattern that provides the data for mathematical reconstruction.