search

PrincetonUniversity / aspire

Algorithms for Single Particle Reconstruction
Other
13 stars 3 forks source link

readme

Logo

Github Releases

ASPIRE - Algorithms for Single Particle Reconstruction

This is the MATLAB version with initial private repository started in 2013. Please check the corresponding Python version started in 2018.

Current version: 0.14 Date: 11/2018

For more information about the project, algorithms, and related publications please refer to the ASPIRE Project website.

Installation

The following steps should be executed only once.

  1. Download and extract the package. We will assume that the package has been extracted to a folder named ASPIRE.
  2. Change into the directory ASPIRE.
  3. Run "initpath".
  4. Run "install".
  5. Change into the directory ASPIRE/projections/class_average/ and run “gen_simulation_data”.

Getting started

  1. Enter the directory ASPIRE.
  2. Run “initpath” This script should be executed each time you start your Matlab session.
  3. Change into the directory ASPIRE/examples
  4. Enjoy the example scripts

Revisions

Changes from version 0.13

  1. Improve speed
  2. Improve reconstruction workflows
  3. Integrate PSWF-based sPCA
  4. Revise class averaging to use PSWF
  5. Add support for EM iterations in class averaging
  6. Add support for the FINUFFT library.
  7. New implementation of Covariance Wiener Filtering (CWF) denosing
  8. Implementation of a 3D fast Fourier-Bessel basis (and a new implementation of the 2D version)
  9. Replace FIRM reconstruction algorithm 10.Add abinitio common-lines algorithms for molecules with Cn symmetry 11.Revise examples 12.Add various utilities

Changes from version 0.12

  1. Added improved denoising for images
  2. Added volume covariance estimation
  3. Added improved noise estimation
  4. Added Fast steerable PCA
  5. New example scripts
  6. Improved ab-initio reconstruction for non-symmetric molecules
  7. Added routines for reconstruction workflow

Changes from version 0.11

  1. Added memory efficient FIRM reconstruction routine reconstruction/FIRM/recon3d_firm_ctf_large.m
  2. Added example file for using cryo_project examples/simulated_projections.m
  3. Update documentation of projections/simulation/cryo_project.m
  4. Remove obsolete benchmark results from comment at the end of ./sinograms/test_commonlines_gaussian.m

Changes from version 0.1

  1. Updated class averaging code.
  2. Revised the function cryo_project to generate projections whose size is different from the size of the projected volume.