Skip to content

Latest commit

 

History

History
99 lines (74 loc) · 5.13 KB

4 RELION Tutorial.md

File metadata and controls

99 lines (74 loc) · 5.13 KB

RELION Tutorial

This section describes the relion tutorial jobs workflow: input parameters, explanation, expected results.

The RELION tutorial consists of 29 jobs of various types. To make our simplified tutorial more understandable, we consider each job as an independent step.

The visualization of the tutorial workflow represented below.

images/relion_tutorial_workflow.png

  • Solid arrows show explicit data interconnectivity between steps. For example, the output data of the CtfFind/ctffind is the input data of the Extract/LoG_based.
  • Bold solid and dashed arrows represents a transition between step chapters.
  • Dashed arrows show implicit data interconnectivity between steps. For example, parameters calculated during the AutoPick/optimize_params step become input values for the AutoPick/template_based step.
  • The first and the last steps marked with the purple color.

Step description structure

Each step description subunit of this simplified RELION tutorial includes the following parts.

  • Navigation block that looks like the example below
Tutorial Unit 1.2 Beam-induced motion correction
Previous Step (1) Import/movies
  List of jobs that will become an input for the current step. ~ prefix means that relation is implicit.
Step Alias (2) MotionCorr/own
  The name for the current task (JobType/alias). It is recommended to set this alias to a completed job of the current step
Next Step (3) CtfFind/ctffind
  List of jobs that will use the output of the current step as input. ~ prefix means that relation is implicit.
  • Recommended parameters table. It looks like the one below. Note that the parameters are similar to the original RELION tutorial at ~95% and some of them differ.
Field name Field value
I/O I/O (Name of tab in parameters input section of RELION GUI)
Input files (Name of parameter) Movies/*.tiff (Value of parameter)
Node type 2D micrograph movies [*.mrcs, *.tiff]
  • Explanation text for the step with pictures and screenshots.

Remember that each step represents a single job only. At some steps, you also need to run terminal applications and Chimera as well.

MPI

Some job types support parallel execution via message passing interface, or MPI. The table below describes the relation of each tutorial step to the MPI and how long a particular job running.

N Name Uses MPI Running time
1 Import/movies No instantly
2 MotionCorr/own No fast
3 CtfFind/ctffind Yes fast
4 ManualPick/illustrate_only No instantly
5 Select/5mics No instantly
6 AutoPick/LoG_based Yes fast
7 Extract/LoG_based Yes fast
8 Class2D/LoG_based Yes slow
9 Select/template4autopick No instantly
10 AutoPick/optimize_params Yes slow
11 AutoPick/template_based Yes average
12 Extract/template_based Yes average
13 Sort/after_autopick No fast
14 Select/after_sort No instantly
15 Class2D/after_sorting Yes slow
16 Select/class2d_aftersort No instantly
17 InitialModel/symC1 Yes fast
18 Class3D/first_exhaustive Yes slow
19 Select/class3d_first_exhaustive No instantly
20 Extract/best3dclass_bigbox Yes average
21 Refine/first3dref Yes slow
22 MaskCreate/first3dref Yes fast
23 PostProcess/first3dref No fast
24 CtfRefine/ctfrefine Yes fast
25 Polish/train No slow
26 Polish/polish No slow
27 Refine3D/polished Yes slow
28 PostProcess/polished No fast
29 LocalRes/polished No fast

Note that:

  • The maximum useful number of MPI processes for all jobs that expect micrograph STAR files as input is the number of micrographs in a file, 1 MPI process per micrograph.
  • MPI workflow in the "3D Refine" job type is 1 master + 2 sets of slaves, so it is recommended to use an odd number of processes that greater than or equal to 3.

All the preparations passed now, and we can proceed with the tutorial.


3 RELION Jobs Top Page 4.1 Prepare Input Data