jspr: software for single particle cryo-EM image processing and 3-D reconstruction

jspr (/ˈjaspər/) is a collection of the programs developed in Wen Jiang group in the Markey Center for Structural Biology, Department of Biological Sciences, Purdue University.

Major Features

  • Support for Icosahedral symmetry (thoroughly tested) and all other point symmetries (tested with ribosome and 20S proteasome, to be more thoroughly tested)
  • Asymmetric reconstruction of structures with mixed symmetries
    • standard asymmetric reconstruction (SAR, i.e. all components are identically arranged in all particles)
    • focused asymmetric reconstruction (FAR, i.e. all components have defined structure but can be arranged differently in different particles)
  • De novo initial model
    • random model approach
    • synthetic shapes (icosahedra and dodecahedra with adjustable size, shell thickness and curvature)
  • Truly independent reconstruction (initial model and refinement) and "gold" standard resolution evaluation
  • Multiple model competition refinements for heterogeneous particles
  • Support merging of multiple datasets imaged using different microscopes or at different magnifications
  • Support small to very large maps (864^3 so far) and datasets (10^5 particles)
  • Refinement and correction of elliptic distortion without the need of pre-calibration of the microscope using polycrystal samples
  • Many filters to choose subset of particles for inclusion in 3-D reconstruction
    • consensus approach: consistent views from multiple independent refinements
    • stable solution approach: stable views from multiple iterations of refinements
    • filtering by scores, defocus, etc. across all micrographs or for individual micrographs
  • Automated CTF fitting
    • support for astigmatic images
    • power spectra fitting approach
    • s2focus approach using graph theory
  • Support for workflow using multiple binning levels (i.e. 4x binning -> 2x binning -> no binning)
  • High resolution refinements (used to solve several near-atomic resolution icosahedral virus structures)
    • grid-less optimization of particle view/position parameters
    • defocus refinement
    • astigmatism refinement and correction
    • magnification refinement
    • anisotropic magnification (elliptic distortion) refinement
  • Parallel processing
    • Computers
      • single desktop/workstation with multi-cores
      • collection of workstations
      • clusters
      • distributed computer nodes pooled using HTCondor
    • Job queues
      • ad-hoc without queues
      • PBS/Torque etc.
      • HTCondors
    • Parallel 3-D reconstruction using MPI with adaptive workload assignment for nodes of different performances
  • Utility tools
    • text-based virtual image format (.lst)
      • create very large datasets from multiple micrographs while keeping the file small
      • flexible operations of datasets: merge, split, common, difference, etc.
      • storage of arbitrary parameters using key=value pairs
      • all these flexible operations without the need to modify the actual binary image files
    • conversion of jspr lst files from/to Relion star files (new in 2017-7-20 snapshot)
    • scanner (Nikon) image format conversion (TIFF to MRC) and pixel value conversion (transmittance to O.D.)
    • euler angle distribution plot in 2-D or 3-D
    • map handedness and symmetry view conversions
    • simulation of density maps from PDB atomic models using CCP4 programs
    • non-crystallographic averaging (NCS)
    • many other utilities

Documentations

References

    This software is built on top of EMAN and EMAN2 with many additions and bug fixes. Please cite the following references in your publication:

Download and Installation

  • Linux x86_64 package: jspr.2017-7-20.tar.gz (598.4 MB)
    New feature: conversion of jspr lst files from/to Relion star files
    Download and then unpack it. All contents will be unpacked into folder jspr. Please follow the instructions in jspr/README to use the software.
  • Linux x86_64 package: jspr.2016-5-4.tar.gz (521.4 MB)
    New feature: elliptic distortion refinement and correction
    Download and then unpack it. All contents will be unpacked into folder jspr. Please follow the instructions in jspr/README to use the software.
  • Linux x86_64 package: cryoem.jspr.2014-02-10.tar.gz (329 MB)
    Download and then unpack it (for example, run tar zxvf cryoem.jspr.2014-02-10.tar.gz in a terminal). All contents will be unpacked into folder cryoem. Please follow the instructions in cryoem/README to use the software.
  • Virtual machine image: cryoem.jiang.2014-02-10.Centos 5.10 x86_64.ova (2.65 GB)
    The image was exported from Virtual Box software (free!) but should be supported by other virtual machine softwares (for example, VMWare, although not tested yet). You can use any host system (Windows, Mac OS X or Linux) to run the Virtual Box software and then import the downloaded image into Virtual Box. The guest OS in this image is CentOS 5.10 x86_64 with root password: cryoem, username/password: cryoem/cryoem. The jspr software was pre-configured for user cryoem and you can run it directly without additional download/installation. You should be able to just open a terminal window and run e2projectmanager.py without any problem. This image also includes a small test dataset (epsilon15 phage) in the directory ~/projects/e15. Follow the Online Tutorial to get started.