BL4-2 Biological Small Angle Scattering/Diffraction

SAXSPipe for Automated SAXS Data Processing and Analysis

Our high-throughput SAXS autosampler can can now collect 96 samples autonomously over a five hour period.  To address the large volume of data this generates we have developed SAXSPipe, an automated SAXS data processing and analysis pipeline.  SAXSPipe requires no user intervention and produces a web viewable table of graphs and statistics that allow the user to rapidly and easily review the results of the SAXS experiments all the way through from basic data quality checks to protein flexibility analysis, oligomer state determination and protein density modeling procedures.   The results of the analysis are stored in an automatic analysis folder that the user can take home and consult later.

The pipeline is written in Python and uses analysis software taken from the ATSAS package by Dimitri Svergun’s group at Hamburg, the SASTBX toolbox headed by Peter Zwart at the LBL, IMP from Andrej Sali’s group at UCSF, and our in-house image scaling and integration software sastool.  The software has been designed to be modular and easily adaptable, allowing the easy addition of new software tools as they arise. SAXSPipe is now regularly used during biological SAXS experiments to help users analyze their data.

SAXSPipe output overview:
SAXSPipe output overview

The output analysis table has the following parts:

    1) Initial data collection information – this section includes on the data collection such as the time it was taken, the file name, uv measurements (if this data is collected), the results of the scaling and integration of the images via sastool, and an rg and distribution function analysis.   Clicking on the .sub file name will bring up the scattering profile in an appropriate program (for instance Primus or Graphit) for further analysis.  The sastool log can also be accessed by clicking on the entries in the ‘buffer(rejected) Sample(rejected)’ column, and the image themselves can be examined by clicking on the ‘images’ link within the same column.
    Intial data collection info


    2) Data collection diagnostic graphs - this section of the table features four graphs to help the user quickly identify problems with the data collection:
    a. The variance of the scaled and integrated images from the buffer and sample data collections, allowing for the rapid identification of radiation damage and other problems like air bubbles
    b. A comparison of the averaged buffer and sample scattering, which allows the user to quickly see how strong the sample scatters
    c. Very low q comparison of the subtracted sample scattering profiles for each image, again allowing the user to identify radiation damage over time
    d. A graph of the change in rg over the image series
    Intial data collection info


    3) Final subtracted file analysis - This section contains a series of basic SAXS data analysis graphs:
    a. q vs log(i) plot - the usual SAXS analysis graph
    b. log(q) vs log(i) plot - As above but with the lower q region emphasized
    c. q2 x I(q) vs q plot(Kratky plot) - Protein flexibility/disorder graph
    Intial data collection info


    4) Automated data analysis graphs - results of automated data analysis:
    a. Autorg results - the results of the automated rg and i0 calculation program from Atsas
    b. Datgnom results - the pr function from Datgnom
    Intial data collection info


    5) Row analysis graphs - comparison of data of all data in a single row:
    a. Buffer file comparison - all buffer in series can be compared to detect changes, possibly indicating capillary fouling
    b. Sub file comparison - quick comparison of all sub files in series
    c. Rg change - a graph of the change in rg across the series
    d. I0 change - a graph of change in i0
    Intial data collection info