Skip to content

3D classification & refinement

Jump to bottom
Michael A. Cianfrocco edited this page Jun 7, 2018 · 1 revision

Home > 3D classification & refinement

##3D classification & refinement In order to analyze and refine 3D structures, you will need a starting model. This model can be downloaded from the PDB or EMDB if the structure is already know. BUT, if the structure is unknown, you will need to use ab initio methods (RCT/OTR) in order to calculate the 3D model in an unbiased manner.

###What program should I use for 3D classification and refinement?

Negative stain:

  • If you have a small negative stain dataset, SPIDER can be used as quick refinement routine that can be run on a single workstation.
  • Alternatively, you can use Relion's auto-refine program to iteratively analyze and reconstruct the single particles relative to the 3D initial model
    • If you are doing low-resolution negative stain EM, you do not need to provide CTF information. Instead you can provide just the particle stack.

If you suspect your sample to be heterogenous, you can also use Relion's 3D classification program to first classify your data into homogenous groups, and then subject the homogenous group to refinement individually.

Cryo-EM:

  • If you have a cryo-EM dataset, you should take advantage of the full spectrum of tools available to you through Relion. This means that you will need to follow the Relion workflow, and you will be able to determine CTF-corrected 3D models of your data.

###3D classification & refinement routines

Relion:

SPIDER:

###Ab initio 3D volume structures using RCT/OTR:

Home

Clone this wiki locally