Problem Dataset

[Nicholas David Clark]

I have a SEC-SAXS dataset which is quite troublesome, due to the odd background intensities (see attached images). I've tried to process the data with REGALS but am unable to correctly deconvolute the signals. 

I'm wondering if anyone has dealt with similar data in the past and has some insights into how to properly proceed with processing? Alternatively, is the data just useless in this state and I'm better off recollecting?

Thanks,

Nick

[Jesse Hopkins]

Hi Nick,

Any time you're asking the question of whether you should recollect data, I tend to say that you should go ahead and do that if possible (especially if you can figure out what was causing the baseline drift and how to get rid of that). That being said, there's a couple of options here, especially for the stronger signal data.

First, you might be able to use a buffer region just before the first small peak, and then do a linear baseline correction across the peaks, just on the downsloping side. If you make the range tight enough, there might not be that much drift, allowing you to get something that might be more reliable.

Second, REGALS can take a lot of playing around with. I'd recommend trying it both with and without subtracting buffer (e.g. fitting the buffer as a component, as well as the drift), and limiting the range for REGALS to where it starts to increase, around frame 215. You might also try cutting off some of the early frames and seeing what affect that has (basically trying to limit the amount of change that you're fitting).

If you send me the dataset I can take a quick look at it when I have time.

All the best.

- Jesse

----

Jesse Hopkins, PhD

Deputy Director

BioCAT, Sector 18

Advanced Photon Source

--
You received this message because you are subscribed to the Google Groups "BioXTAS RAW" group.
To unsubscribe from this group and stop receiving emails from it, send an email to bioxtas_raw...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/bioxtas_raw/cd9cd032-7e69-44e0-8605-7917045457cdn%40googlegroups.com.

[Nicholas David Clark]

Hi Jesse,

Thanks for the recommendations and the offer of assistance! I've sent you the dataset off-list.

Best,

Nick

[Jesse Hopkins]

Just to close the loop on this question, we determined that the data hadn’t been properly normalized. After applying the correct normalization, the integrated intensity looks like you’d expect for a SEC-SAXS dataset.

All the best.

- Jesse

----

Jesse Hopkins, PhD

Deputy Director

BioCAT, Sector 18

Advanced Photon Source

To view this discussion on the web visit https://groups.google.com/d/msgid/bioxtas_raw/e4b6e05c-fa7f-4f6d-84bc-8f28fdcd0342n%40googlegroups.com.

[Nicholas David Clark]

Should anyone run into a similar issue (Panel A), Jesse figured out that the main issue was incorrect scaling of the water peak (normalization) during the series. We determined this by decomposing the series into the component profiles, at which time it became obvious that the scaling was poor (Panel B). We scaled to the water peak in RAW (Panel D) and then exported the correctly scaled profiles. We then replotted the series in RAW from the exports and the series now should be usable (Panel C).