Emperor plot stuck with 550 samples

User Support
1

Hi,

I have tested the emperor plot demo in Moving Pictures and I can see all the plots(33 samples).

47%20AM
47%20AM1060Ă—808 44 KB

However, the emperor plot would stop at the beginning when I was trying to plot 550 samples.
11%20AM
11%20AM1264Ă—770 21 KB

Could anyone tell me what I should do?

Thanks,

Hongwei

2

Hi @katon,
This sounds like a memory issue on your computer — emperor will start just fine with 33 samples but 550 cause your browser to hang when loading this visualization. You could try opening on a different computer, or just wait (it should eventually open unless if it is causing your browser to crash).

Good luck!

3

Thank you for the help, Nicholas!

1 Like
4

Hi Nicholas,

I have tried it on a Dell Alienware Aurora still it doesn’t work.

Dell Alienware Aurora’s configuration is:
Intel® Core™ i7-7700
32GB 2400MHz DDR4
256GB (SSD),
GeForce GTX 1080 Ti

Now I feel it should be emperor’s problem instead of computer’s problem.

6

Hi @katon,

Would you be able to share the artifact with me? Alternatively if you could you copy the output printed on the JavaScript console when Emperor fails to load?

To open the JavaScript console in Google’s Chrome go to View > Developer > JavaScript Console.

8

Hi Yoshiki,

It seems there are errors caused by different metadata rows number when the emperor is loading files.
30%20PM

Also I have checked the log file while generating core-metrics. There are warnings like:

RuntimeWarning: The result contains negative eigenvalues. Please compare their magnitude with the magnitude of some of the largest positive eigenvalues. If the negative ones are smaller, it's probably safe to ignore them, but if they are large in magnitude, the results won't be useful. See the Notes section for more details. The smallest eigenvalue is -0.34533786907859787 and the largest is 17.217787970350557.

Do you think these warnings are related to those different metadata rows numbers?

Thanks,

Hongwei

10

@katon, I think this might have to do something with your metadata. Have you been able to use your mapping file with another plugin? For example with qiime feature-table summarize? Otherwise, I think I would need to have a look at your metadata.

12

Hi Yoshiki,

Sorry for the late reply. I just check the metadata file I have been using. It is actually quite simple. Just one column with column name id and column content are sample names.

Yes, I have used other plugin like the qiime feature-table summarize and it did work and generate some results.

Do you think the only possible reason might due to the metadata file?

Thanks,

Hongwei

13

Hi @katon, If you can share the metadata and the ordination with me, that would be very useful. Otherwise I am a bit of a loss here on how else I can help.

It sounds like this is a bug with emperor. :confused:

14

Hi Yoshiki,

Please check the attached metadata and the ordination file(weighted_unifrac_emperor.qzv).
metadata.tsv (12.3 KB)
weighted_unifrac_emperor.qzv (813.9 KB)

Thanks!

Hongwei

1 Like
15

@katon, just a quick thing, the ordination file will be something like weighted_unifrac_pcoa_results.qza (qiime tools peek should say the type is PCoAResults ), you’ve uploaded the visualization.

17

Thanks for pointing it out, Evan!

Here it is:
weighted_unifrac_pcoa_results.qza (1.1 MB)
weighted_unifrac_distance_matrix.qza (2.3 MB)

1 Like
18

Hi @katon, thanks for posting these artifacts. It looks like you've found a bug in Emperor where the metadata needs to have at least one ore column (in addition to the sample identifier). I've generated a metadata file with an extra column and things seem to work just fine:

metadata-fixed.tsv (15.5 KB)

As a side note, I've added a ticket to Emperor's issue tracker:

3 Likes
20

Thanks a lot for your help, Yoshiki!

Hongwei