Back to PLS Help

the opposite structural covarience of the same clusters in different LVs

I'm Online
guoo
Posted on 08/19/17 23:07:50
Number of posts: 10
guoo posts:

Greetings PLS experts,


I recently ran a seed PLS analysis focused on the structural covariance of the anterior and posterior hippocampus. I followed the Persson and Spreng (2014) paper "Sex differences in volume and structural covariance of the anterior and posterior hippocampus", in the sense that I used the volume values of anterior and posterior hippocampus as the values of seeds and run a 'Regular Behav PLS'. however, I am so confused with information regarding the interpretation of results in the paper.

First of all, in the section “Structural covariance of the hippocampus” in the paper, the clusters with positve covariance or negative covariance described in the paper, only cover part of Table 2,while some clusters inclued in Table 2 are not mentioned, both in LV1 and LV2.  Is there a standard to determine which clusters should be described?

Secondly, acoording to Fig.1 and Table 2 in the paper, I can get converse relation between the same brain regions and the seed regions. For example, In LV1, the superior parietal lobule (R) shows the negative covariance with the aHC and pHC in both men and women, however it shows the positive covariance with the pHC in both men and women in LV2. so, which one(positive/negative) should be the actual relation between superior parietal lobule (R) and pHC? If both are correct, how to interpret it? Also in LV2, the middle temporal gyrus (R) showed both positive and negative saliences, How to interpret this correctly?

I am so sorry for this very basic question. Thanks in advance for any advice.

guo o

Replies:

Untitled Post
rmcintosh
Posted on 08/20/17 05:33:30
Number of posts: 394
rmcintosh replies:

Hi Guo - thanks for tyour question.   Please see my replies below

"First of all, in the section “Structural covariance of the hippocampus” in the paper, the clusters with positve covariance or negative covariance described in the paper, only cover part of Table 2,while some clusters inclued in Table 2 are not mentioned, both in LV1 and LV2.  Is there a standard to determine which clusters should be described?"
 
Not really.  We don't do statistics on the cluster sizes so its up to the user to ensure the cluster report in table adequately reflects the main areas in the efftect.  
 
 
"Secondly, acoording to Fig.1 and Table 2 in the paper, I can get converse relation between the same brain regions and the seed regions. For example, In LV1, the superior parietal lobule (R) shows the negative covariance with the aHC and pHC in both men and women, however it shows the positive covariance with the pHC in both men and women in LV2. so, which one(positive/negative) should be the actual relation between superior parietal lobule (R) and pHC? If both are correct, how to interpret it? Also in LV2, the middle temporal gyrus (R) showed both positive and negative saliences, How to interpret this correctly?"
 
Keep in mind the seed correlations is only 1/2 of the output from PLS.  The other part (fig2 and 3), show different spatial patterns, suggesting the regions that show that correlation with seed are distinct between LV's.  Also in the table, you should note that teh middle temporal max reported are spatially distinct.
 


Untitled Post

I'm Online
guoo
Posted on 08/20/17 10:35:34
Number of posts: 10
guoo replies:

quote:

Hi Guo - thanks for tyour question.   Please see my replies below

"First of all, in the section “Structural covariance of the hippocampus” in the paper, the clusters with positve covariance or negative covariance described in the paper, only cover part of Table 2,while some clusters inclued in Table 2 are not mentioned, both in LV1 and LV2.  Is there a standard to determine which clusters should be described?"
 
Not really.  We don't do statistics on the cluster sizes so its up to the user to ensure the cluster report in table adequately reflects the main areas in the efftect.  
 
 
"Secondly, acoording to Fig.1 and Table 2 in the paper, I can get converse relation between the same brain regions and the seed regions. For example, In LV1, the superior parietal lobule (R) shows the negative covariance with the aHC and pHC in both men and women, however it shows the positive covariance with the pHC in both men and women in LV2. so, which one(positive/negative) should be the actual relation between superior parietal lobule (R) and pHC? If both are correct, how to interpret it? Also in LV2, the middle temporal gyrus (R) showed both positive and negative saliences, How to interpret this correctly?"
 
Keep in mind the seed correlations is only 1/2 of the output from PLS.  The other part (fig2 and 3), show different spatial patterns, suggesting the regions that show that correlation with seed are distinct between LV's.  Also in the table, you should note that teh middle temporal max reported are spatially distinct.
 

Hi Randy,
Thank you very much for your quick reply.

For your reply, I have two questions. Firstly, we can see that clusters contained in Table2 are all exceeding 50 voxels with BSR>3.0 in the paper, However, the author only selected a part of these clusters to discuss in the "Discussions" section. I doubt whether this is not comprehensive enough.

Secondly, superior parietal lobule (R) shows the negative covariance with the aHC and pHC in both men and women in LV1, can I understand that the larger the aHC and pHC, the smaller the volume of superior parietal lobule (R), in both men and women? however, the superior parietal lobule (R) shows the positive covariance with the pHC in both men and women in LV2, can I understand that the larger the pHC, the larger the volume of superior parietal lobule (R)? Obviously, these two conclusions are contradictory. The clusters of hippocampus have the same conditions, the left and right hippocampus shows the positive covariance with the aHC and pHC in both men and women in LV1,while the  left and right posterior hippocampus shows the negative covariance with the aHC in both men and women, these two conclusions are also contradictory. Can these seemingly contradictory results also be attributed to the spatial differences in clusters, even though the clusters belong to the same brain region?

I am afraid that I have a wrong understanding of the conception and difference of different LVs.

Thanks again for any help.

guo o



Untitled Post
rmcintosh
Posted on 08/21/17 20:57:23
Number of posts: 394
rmcintosh replies:

quote:

Hi Randy,
Thank you very much for your quick reply.

For your reply, I have two questions. Firstly, we can see that clusters contained in Table2 are all exceeding 50 voxels with BSR>3.0 in the paper, However, the author only selected a part of these clusters to discuss in the "Discussions" section. I doubt whether this is not comprehensive enough.

Secondly, superior parietal lobule (R) shows the negative covariance with the aHC and pHC in both men and women in LV1, can I understand that the larger the aHC and pHC, the smaller the volume of superior parietal lobule (R), in both men and women? however, the superior parietal lobule (R) shows the positive covariance with the pHC in both men and women in LV2, can I understand that the larger the pHC, the larger the volume of superior parietal lobule (R)? Obviously, these two conclusions are contradictory. The clusters of hippocampus have the same conditions, the left and right hippocampus shows the positive covariance with the aHC and pHC in both men and women in LV1,while the  left and right posterior hippocampus shows the negative covariance with the aHC in both men and women, these two conclusions are also contradictory. Can these seemingly contradictory results also be attributed to the spatial differences in clusters, even though the clusters belong to the same brain region?

I am afraid that I have a wrong understanding of the conception and difference of different LVs.

Thanks again for any help.

guo o

Hi Guo - its worth keeping in mind that regions need not be homogeneous in their responses.  Large regions of interest may have small parts that relate differently.

I might suggest that if you have other questions about the paper itself, you should contact the authors directly as I was not involved in this work.



Untitled Post

I'm Online
guoo
Posted on 08/22/17 01:32:13
Number of posts: 10
guoo replies:

Thank you so much for the help randy, really appreciate it.

guo o




Login to reply to this topic.

  • Keep in touch

Enter your email above to receive electronic messages from Baycrest, including invitations to programs and events, newsletters, updates and other communications.
You can unsubscribe at any time.
Please refer to our Privacy Policy or contact us for more details.

  • Follow us on social
  • Facebook
  • Instagram
  • Linkedin
  • Pinterest
  • Twitter
  • YouTube

Contact Us:

3560 Bathurst Street
Toronto, Ontario
Canada M6A 2E1
Phone: (416) 785-2500

Baycrest is an academic health sciences centre fully affiliated with the University of Toronto