# Localize.pytom: a modern webserver for cryo-electron tomography

## TL;DR

Localize.pytom is a free webserver that helps identify and locate macromolecules in cryo-electron tomograms for structural analysis.

## Contribution

It introduces a modern web-based platform with tools for accurate localization and visualization of macromolecules in cryo-electron tomography data.

## Key findings

- The server provides coordinates and densities of candidate macromolecules in tomograms.
- Users can choose between coarse or fine angular sampling strategies to improve localization accuracy.
- A novel web-density viewer allows further analysis of macromolecular averages and score distributions.

## Abstract

Localize.pytom, available through http://localize.pytom.org is a webserver for the localize module in the PyTom package. It is a free website and open to all users and there is no login requirement. The server accepts tomograms as they are imaged and reconstructed by Cryo-Electron Tomography (CET) and returns densities and coordinates of candidate-macromolecules in the tomogram. Localization of macromolecules in cryo-electron tomograms is one of the key procedures to unravel structural features of imaged macromolecules. Positions of localized molecules are further used for structural analysis by single particle procedures such as fine alignment, averaging and classification. Accurate localization can be furthermore used to generate molecular atlases of whole cells. Localization uses a cross-correlation-based score and requires a reference volume as input. A reference can either be a previously detected macromolecular structure or extrapolated on the server from a specific PDB chain. Users have the option to use either coarse or fine angular sampling strategies based on uniformly distributed rotations and to accurately compensate for the CET common ‘Missing Wedge’ artefact during sampling. After completion, all candidate macromolecules cut out from the tomogram are available for download. Their coordinates are stored and available in XML format, which can be easily integrated into successive analysis steps in other software. A pre-computed average of the first one hundred macromolecules is also available for immediate download, and the user has the option to further analyse the average, based on the detected score distribution in a novel web-density viewer.