| 1 |
Advanced Series processing – Singular value decomposition (SVD) |
| 2 |
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ |
| 3 |
.. _raw_svd: |
| 4 |
|
| 5 |
Sometimes SEC fails to fully separate out different species, and you end up with overlapping |
| 6 |
peaks in your SEC-SAXS curve. It is possible to apply more advanced mathematical techniques |
| 7 |
to determine if there are multiple species of macromolecule in a SEC-SAXS peak, and to attempt |
| 8 |
to extract out scattering profiles for each component in an overlapping peak. Singular value |
| 9 |
decomposition (SVD) can be used to help determine how many distinct scatterers are in a |
| 10 |
SEC-SAXS peak. :ref:`Evolving factor analysis (EFA) <raw_efa>` is an extension |
| 11 |
of SVD that can extract individual components from overlapping SEC-SAXS peaks. |
| 12 |
Note that the first step of EFA is doing SVD, but that happens entirely within |
| 13 |
the EFA analysis window. The SVD window does not need to be opened before doing |
| 14 |
EFA. This tutorial covers SVD. |
| 15 |
|
| 16 |
A video version of this tutorial is available: |
| 17 |
|
| 18 |
.. raw:: html |
| 19 |
|
| 20 |
<style>.embed-container { position: relative; padding-bottom: 56.25%; height: 0; overflow: hidden; max-width: 100%; } .embed-container iframe, .embed-container object, .embed-container embed { position: absolute; top: 0; left: 0; width: 100%; height: 100%; }</style><div class='embed-container'><iframe src='https://www.youtube.com/embed/Hyvr18s7Ihg' frameborder='0' allowfullscreen></iframe></div> |
| 21 |
|
| 22 |
The written version of the tutorial follows. |
| 23 |
|
| 24 |
|
| 25 |
#. Clear all of the data in RAW. Load the **phehc_sec.hdf5** file in the **series_data** folder. |
| 26 |
|
| 27 |
* *Note:* The data were provided by the Ando group at Cornell University |
| 28 |
and is some of the data used in the paper: *Domain Movements upon Activation of |
| 29 |
Phenylalanine Hydroxylase Characterized by Crystallography and Chromatography-Coupled |
| 30 |
Small-Angle X-ray Scattering*\ . Steve P. Meisburger, Alexander B. Taylor, Crystal |
| 31 |
A. Khan, Shengnan Zhang, Paul F. Fitzpatrick, and Nozomi Ando. Journal of the |
| 32 |
American Chemical Society 2016 138 (20), 6506-6516. `DOI: 10.1021/jacs.6b01563 |
| 33 |
<https://dx.doi.org/10.1021/jacs.6b01563>`_ |
| 34 |
|
| 35 |
|efa_series_plot_png| |
| 36 |
|
| 37 |
#. Right click on the **phehc_sec.hdf5** item in the Series list. Select the “SVD” option. |
| 38 |
|
| 39 |
#. The SVD window will be displayed. On the left are controls, on the right are plots of |
| 40 |
the value of the singular values and the first autocorrelation of the left and right |
| 41 |
singular vectors. |
| 42 |
|
| 43 |
* *Note:* Large singular values indicate significant components. What matters is the relative |
| 44 |
magnitude, that is, whether the value is large relative to the mostly flat/unchanging |
| 45 |
value of high index singular values. |
| 46 |
|
| 47 |
* *Note:* A large autocorrelation indicates that the singular vector is varying smoothly, |
| 48 |
while a low autocorrelation indicates the vector is very noisy. Vectors corresponding to |
| 49 |
significant components will tend to have autocorrelations near 1 (roughly, >0.6-0.7) and |
| 50 |
vectors corresponding to insignificant components will tend to have autocorrelations near 0. |
| 51 |
|
| 52 |
|svd_panel_png| |
| 53 |
|
| 54 |
#. Adjust the starting frame number to 100, the ending frame number to near 300, and switch |
| 55 |
to using Subtracted data. |
| 56 |
|
| 57 |
* *Note:* The blue points are in the plot on the left are the region being |
| 58 |
used for SVD, while the red points show the rest of the SEC-SAXS curve. |
| 59 |
|
| 60 |
|svd_panel_sub_png| |
| 61 |
|
| 62 |
#. We have now isolated the peak. Looking at the top plot, we see there are two singular |
| 63 |
values significantly above the baseline level, and from the autocorrelation we see two |
| 64 |
values with both left and right singular vectors autocorrelations near 1. This indicates |
| 65 |
that there are two scattering components in the peak, even though there are no obvious |
| 66 |
shoulders in the region we selected |
| 67 |
|
| 68 |
* *Try:* Adjust the starting and ending values and seeing how that changes the SVD |
| 69 |
results. Is there a region of the peak you can isolate that has just one significant |
| 70 |
component? |
| 71 |
|
| 72 |
* *Note:* Normally, changing between Unsubtracted and Subtracted SEC-SAXS profiles |
| 73 |
should remove one significant singular value component, corresponding to the buffer |
| 74 |
scattering. In this data, you will see no difference, as the profiles used to |
| 75 |
produce the SEC-SAXS curve were already background subtracted. |
| 76 |
|
| 77 |
* *Note:* You can save the SVD plots by clicking the Save button, as with the plots |
| 78 |
in the main RAW window. You can save the SVD results, either just the plotted values |
| 79 |
or all of the values, using the two Save buttons in the SVD panel. |
| 80 |
|
| 81 |
|singular_values_png| |
| 82 |
|
| 83 |
#. Close the SVD window by clicking the OK button. |
| 84 |
|
| 85 |
|
| 86 |
Note: By default, RAW bins the profiles before doing an SVD, in order to speed |
| 87 |
up the actual SVD. You can turn binning on and off and adjust the binning |
| 88 |
parameters in the Series options panel in the Advanced Options window. |
| 89 |
|
| 90 |
.. |efa_series_plot_png| image:: images/efa_series_plot.png |
| 91 |
:target: ../_images/efa_series_plot.png |
| 92 |
|
| 93 |
.. |svd_panel_png| image:: images/svd_panel.png |
| 94 |
:target: ../_images/svd_panel.png |
| 95 |
|
| 96 |
.. |svd_panel_sub_png| image:: images/svd_panel_sub.png |
| 97 |
:target: ../_images/svd_panel_sub.png |
| 98 |
|
| 99 |
.. |singular_values_png| image:: images/singular_values.png |
| 100 |
:width: 400 px |
| 101 |
:target: ../_images/singular_values.png |