Bioisis

ABOUT SCATTER

ScÅtter is a JAVA-based application for basic analysis of SAXS datasets. ScÅtter is developed by Robert Rambo at the Diamond Light Source (Didcot, UK). The program was previously developed at the SIBYLS beamline (12.3.1) of the Advanced Light Source, Berkeley, CA.

Current Release 3.0j - 2017-03-27 00:00:03 UTC

Download via BioIsis

  • JAVA requirement:

    JAVA 8 (platform 1.8) or greater for Windows, Mac OS X or LINUX

  • Starting ScÅtter:

    ScÅtter comes as a compressed zip file. Unzip the file into a directory, typically accomplished by double clicking the downloaded file. If the file was unzipped in a "Downloads" directory, simply drag the entire unzipped directory to a more permanent location.

    Scatter_directory

    Open the directory and double-click "scatter.jar". The *.jar file can be moved to any suitable directory, just be sure to have the directory available to the system by setting appropriate path statements.

    Windows and LINUX users! Depending on security settings, ScÅtter may be started from command line using "java -jar scatter.jar"

    Apple OSX users: Scatter can be added to the dock. Just drag the jar file to the area next to the trash can. See link.

  • Intensity file:

    ScÅtter can read any 3 column text file (column 1: q, column 2: Intensity, column 3: error ) delimited by space or tabs. Header or footer information starting with # will be ignored. In addition, ScÅtter can read Bruker *.brml SAXS file formats. If you are experiencing problems, please contact Robert Rambo.

Future Updates

  • Independent component analysis for resolving individual scattering curves of components in a mixture.
  • Direct fitting of PDB models to scattering profiles.

Disclaimer

JAVA is a platform independent language that is entirely objected oriented. We tried to optimize the analysis around the object oriented platform. ScÅtter is currently not released as an open source package but we are willing to share source code with collaborators.


Change Log

Release 3.0j - 27 March 2017

minor update:

  • Ignores intensity values that are recorded as zero, sometimes encountered in how the images are masked

Release 3.0k - 11 March 2017

Minor Update:

  • Right click on loaded data files in Analysis tab, can now access some functions.
  • Fixes bug in conversion from nm to Angstroms in Subtract Panel
  • Added Form Factor fitting (details to come)
  • Last version with DAMMIN, IKETAMA (BUBBLES) has finished

Release 3.0i - 20 October 2016

Minor Update :

  • Select Deselect files form analysis tab by highlight with mouse and right click
  • Export data from Dimenionless Kratky and QIQ plot

Release 3.0h - 04 October 2016

Changes to Subtract Tab:

  • Signal Plot can be calculated using Subtracted Data (just add subtracted set into Samples List and hit Signal Plot).
  • Can exclude points in Auto-Rg to reduce noise near beamstop
  • Threshold is saved to config file

Release 3.0g - 22 September 2016

General International version:

  • Spanish, Dannish and Finnish users were having significant delays in loading files from Subtract. This update fixes the problem and should be a widely compatible version.

Release 3.0f - 22 September 2016

Critical Update:

  • Scatter has been updated, the inverse Fourier transform has been updated to standardised the datasets first then perform the transform. This works much better with the smoothness constraint and the chi2 and sk2 parameters identify proper fits better. Also, the error estimates from the model have been updated using a Monte Carlo sampling method (publication soon).
  • A new SVD button has been added to Analysis and Subtraction tab. You can use this to determine the complexity of the profiles prior to merging.

Release 3.0d - 25 June 2016

Save Feature in Subtract:

  • After loading files such as HPLC data, you can save the set to a specified directory or if you are merging frames and have select buffers in the Buffers List, you can save the entire set into a specified directory.

Release 3.0c - 09 June 2016

Scatter Version 3.0 Released:

  • Scatter has been completely rebuilt and optimised with many new features.
  • Automated trace calculation for HPLC SEC SAXS
  • Internal loading of background files and peak selection from loaded set
  • SVD based Averaging
  • Multiple dataset similarity comparison tool
  • Real-Space Transformation Algorithm has been improved and minimises on smoothness
  • Guinier Peak Analysis (New method proposed by Chris Putnam at Univ. of San Diego)
  • Drag-n-Drop PDB files directly into Files Panel
  • PDB file will be converted to a P(r)-distribution and Intensity
  • Archiving : Checked datasets will be consolidated into specified directory along with image files
  • Requirements:
  • JAVA 1.7 or better.
  • On MACS, if using ATSAS, Scatter will not work with the latest release, must use prior release
  • On Windows, if using 64-bit machine, please install a 64-bit JAVA

Release 3.0c - 09 June 2016

Scatter Version 3:

  • Dear Users,
  • Scatter has been completely rebuilt and optimised with many new features.
  • New Features:
  • HPLC SAXS data set analysis
  • -automated trace calculation
  • -background setting and peak selection
  • -SVD averaging
  • Multiple dataset similarity comparison tool
  • Real-Space Transformation
  • -Algorithm has been improved and minimises on smoothness
  • Guinier Peak Analysis
  • -New method proposed by Chris Putnam at Univ. of San Diego
  • -Excellent method for assessing sample quality
  • Drag-n-Drop PDB files directly into Files Panel
  • -PDB file will be converted to a P(r)-distribution and Intensity
  • Archiving
  • -Checked datasets will be consolidated into specified directory along with image files
  • Requirements:
  • JAVA 1.7 or better.
  • On MACS, if using ATSAS, Scatter will not work with the latest release, must use prior release
  • On Windows, if using 64-bit machine, please install a 64-bit JAVA
  • If you have any questions, please feel free to email me.
  • Best,
  • Rob

Release 2.3h - 11 April 2015

Added SVD functionality:

  • Added SVD algorithm to Subtraction tab that allows SVD analysis of the set of frames
  • SVD averaging is a noise reduction technique that reconstructs dataset using only one singular value
  • Averaging now uses a weighted average scheme
  • PDF writer coming soon

Release 2.3f - 29 January 2015

Minor Bug Fixes with new file copy feature:

  • Truncation of datasets by q-value is faster
  • Fixed bug that would delay clicking on or off files (check box)
  • Provide auto output of signal plot from buffer subtraction tab
  • Formatted output of P(r) calculation from PDB (output is directed in Messages tab)
  • Changes to file description in Analysis Tab creates a copy of original file with new name

Release 2.3a - 27 October 2014

Critical Fix and Improved P(r) fits with proper Chi2 and second derivative tests:

  • Added title notes to each file panel
  • Chi2 is properly calculated during P(r) fits, best fit is near 1.0
  • Added kurtosis and second derivative sum to P(r) fit, best fit is near 0
  • Refinement of P(r) automatically writes our refined dataset and runs autoGnom
  • Added new StatMethods class
  • Fixed plotting error in Vc plot
  • Fixed indexing error during merging of datasets in averaging and median by converting containers to HashSets
  • Fixed main scatter plot, uses SingletonClass to maintain a single instance ("toggle crosshairs" doesn't grow in drop down)

Release 2.2b - 04 September 2014

Update fixes critical accumulation error and adds new P(r) features:

  • Fixed critical floating point calculation in the averaging function (subtraction tab).
  • Fixed anti-log scaling in P(r) function
  • Modified autoRg for speed
  • q_min or q_max can be set for entire datasets by clicking on column header in Analysis tab
  • PDB file can be uploaded for direct P(r) calculation in P(r) tab
  • Auto-determine dmax from PDB file
  • Added new plot to Subtraction tab (for SEC SAXS)
  • Output directory in Subtraction tab can now be set independent of main directory
  • For specified datasets, Scatter will auto convert nm-1 to Angstrom-1

Release 2.1 c - June 30, 2014

Version 2.1c fixes a few bugs and has some added features by request:

  • Data is exported on Results tab
  • Filenames are no longer auto-generated (2FILE P(r) tab)
  • The previous specified working directory will be set on start-up
  • Limited access to ATSAS (DATGNOM AND DAMMIN/F) from ScÅtter GUI
  • User must specify directory of the ATSAS binaries in Settings tab
  • 2File (on P(r) tab) automatically calls DATGNOM on the data visible in plot
  • Output from DATGNOM is auto-loaded into DAMMIN/F tab for modeling.
  • Truncation of data in manual Guinier analysis carries over to the working dataset in Analysis tab
  • Crosshairs can be toggled off by right-click on graph (popup menu)

Release 2.0 beta - January 3, 2014

Version 2.0 is a major re-write of the entire program. Due to several bugs and user feedback, Scatter was refactored and improved with the following additional features:

  • Unlimited data files per collection
  • Output SVG and PDF formats
  • Colors and point size editor for individual datasets
  • Buffer subtraction
  • Multi-file chifree calculations
  • Regularized inverse transform using L1-norm
  • Complex formation detection algorithm
  • I(0)/Rg plots over all visible datasets

This is a beta release and we appreciate bug reporting from our users.

Release 1.7i - July 31, 2013

Fixed bug in windows 7, 8 and LINUX, program would not start due to error in determining home directory.

Release 1.7f - July 20, 2013

Fixed Windows 8 bug (directory determination would halt start of program).

Corrected scaling of mass when scale factor is not 1.0 when using Vc button.

Release 1.7e - July 19, 2013

Fixed repopulating of Results tab.

Release 1.7d - July 18, 2013

Fixed auto-zoom that was only affecting data with the max I(q) < 1.

Fixed bug in Guinier fitting. For datasets with I(q) < 1 in Guinier region, auto_Rg and manual Guinier would initially report -1 and 0 for I(0) and Rg until data was adjusted manually. Now, Scatter reports the proper parameters.


Release 1.7c - July 16, 2013

Fixed Chi_free calculation. FOXS rounds the q-value to one less significant figure. Scatter now actively rounds the number to find corresponding number in FOXS fit file. Rounded numbers are printed to console.