Gingko Web Platform: Copy number calls stalling with FACS input.

[andrewrlynch]

I'm analyzing a set of about 300 single cells on the Gingko web app. The karyotypes are very strange so I input a table of predicted ploidy for the cells by DAPI content. The problem is that the data processing is stalling out at certain cells. I can exclude these cells and obviously it will skip them, however it will stall out again on a different cell. Some example ploidy values that this has occurred on are 1.3, 3.4 and 3.9. Any ideas? Thanks!

[mschatz]

Can you send the link to the results that were failing? We also had some issues on the server the other day that might have caused some things to fail

THanks

Mike

On Mon, Nov 19, 2018 at 7:41 PM Andrew Lynch [email protected] wrote:

I'm analyzing a set of about 300 single cells on the Gingko web app. The karyotypes are very strange so I input a table of predicted ploidy for the cells by DAPI content. The problem is that the data processing is stalling out at certain cells. I can exclude these cells and obviously it will skip them, however it will stall out again on a different cell. Some example ploidy values that this has occurred on are 1.3, 3.4 and 3.9. Any ideas? Thanks!

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/robertaboukhalil/ginkgo/issues/13, or mute the thread https://github.com/notifications/unsubscribe-auth/AAL985YdkuQafGIp3xKK53ga8o7CursBks5uw0-5gaJpZM4YqKI8 .

[andrewrlynch]

Thanks for the quick reply. Here is the link: http://qb.cshl.edu/ginkgo?q=results/6VZust8tj6mK3la8lzxn

[jpritt]

Hi Andrew, I am currently working on improving Ginkgo so I took a look at your problem. It appears the facs file was responsible for stalling, as when I ran the analysis without the file it finished quickly. However I don't see your facs file in the directory - would you mind uploading it here or adding it to the project so I can take a look?

Thanks, Jacob

[andrewrlynch]

Hi Jacob,

Thanks for looking into this. I've attached the FACS file. I've also noticed that when I removed the FACS file, the analysis completes. However, the predicted ploidy from the analysis without the FACS data does not line up with the FACS data itself, so I wanted to include this in the analysis. I appreciate the assistance on this.

All the best, Andrew

DerivedIndexPloidies_CombinedSubmitted_NormToRPE1Tri12.txt

[robertaboukhalil]

Hi all,

After running a few tests, it seems the pattern is that if the ploidy in the FACS file is < 1.5, the cell fails; otherwise, it goes through successfully.

Looking at process.R:L203, we currently look for ploidy within the range 1.5 to 6 in increments of 0.05, so if the FACS ploidy file has numbers outside the range, Ginkgo will fail.

I suppose we could parse the user ploidy file and use the numbers specified by the user instead of the fixed range.

[andrewrlynch]

Hi Robert,

Thanks for looking into this. Any recommendations for next steps on my end? Also, out of curiosity, what is the utility of having the lower limit of ploidy at 1.5?

Andrew

[mschatz]

Id vote for a user option for the minimum ploidy. This could be very useful, for example, if analyzing sperm that have a haploid genome and would need to set the minimum to 1.0 or a bit less.

Thanks!

Mike

On Wed, Nov 21, 2018 at 9:23 PM Andrew Lynch [email protected] wrote:

Hi Robert,

Thanks for looking into this. Any recommendations for next steps on my end? Also, out of curiosity, what is the utility of having the lower limit of ploidy at 1.5?

Andrew

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/robertaboukhalil/ginkgo/issues/13#issuecomment-440893051, or mute the thread https://github.com/notifications/unsubscribe-auth/AAL9847LtypKEYOgDWkaQ0vTfFlOIW-4ks5uxgqsgaJpZM4YqKI8 .

[andrewrlynch]

I think the option to set a minimal ploidy would be great, or at least, in the case of attaching the FACS estimated ploidy, pulling the lowest estimated value.

Andrew

[robertaboukhalil]

Hi Andrew, sounds good, I'll take a look at adding that in and get back to you soon

[andrewrlynch]

Hi Folks,

Thanks for working on this. Has this been deployed to the web platform? It looks like the minimum ploidy is still 1.5 and I have been having the same issues when running the data files I shared here in conjunction with the FACS-estimated ploidy values.

All the best, Andrew

[robertaboukhalil]

cc @mschatz

[andrewrlynch]

Hi @mschatz and @robertaboukhalil,

I hope this is okay. I appreciate you both putting the time in on this. I'm just curious if this has been implemented for the web app? I have flow data I would like to incorporate into my analysis, but it looks like it is still timing out on some of the files as I mentioned previously. It also looks like the minimum ploidy is 1.5. Would it be possible to lower this to 1 or lower? We have highly variable samples due to chromosome missegregation.

All the best, Andrew