Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

obtaining integer CN estimates from non-diploid chromosomes #208

Open
7ehuang opened this issue Jan 25, 2024 · 3 comments
Open

obtaining integer CN estimates from non-diploid chromosomes #208

7ehuang opened this issue Jan 25, 2024 · 3 comments

Comments

@7ehuang
Copy link

7ehuang commented Jan 25, 2024

Thank you for the active maintenance of this software! I read in issue #81 and #88 that I can obtain copy number estimates by multiplying the "CNV level" column by 2 and rounding to the nearest integer. However, would this still be the case when analyzing data from cells with aneuploidy (i.e. K562, which is predominantly triploid)? Because then in these cases, the majority of the genome is not necessarily diploid.

I ask because I am comparing my CNVpytor calls with the results from this paper, and the CN calls are highly discordant.

@arpanda
Copy link
Member

arpanda commented Feb 5, 2024

We have an example for the K562 in the cnvpytor manuscript (supplymentary figure S1). In that figure, the normalization is applied based on the bimodal hypothesis (sum of two gaussion distributions with ratio of mean values equal to 2, related to copy numbers 2 and 4 ).

The implementation is hard coded and let you know once we add a user option to enable it through command line.

If you want to unlock option by changing the code, you can remove False and in line of file root.py, under implementation of the caller function you are using :

bim = fit_bimodal(bins[:-1], hist)
if False and bim is not None:

Thank you,
Arijit

7ehuang added a commit to 7ehuang/CNVpytor that referenced this issue Feb 5, 2024
@7ehuang
Copy link
Author

7ehuang commented Feb 5, 2024

We have an example for the K562 in the cnvpytor manuscript (supplymentary figure S1). In that figure, the normalization is applied based on the bimodal hypothesis (sum of two gaussion distributions with ratio of mean values equal to 2, related to copy numbers 2 and 4 ).

The implementation is hard coded and let you know once we add a user option to enable it through command line.

If you want to unlock option by changing the code, you can remove False and in line of file root.py, under implementation of the caller function you are using :

bim = fit_bimodal(bins[:-1], hist)
if False and bim is not None:

Thank you, Arijit

Thank you for your reply! Does this mean that by default (i.e. without making the changes to the code that you suggested), the normalization is not based on the bimodal hypothesis?

Just to clarify, are the results from the manuscript based on the default mean-shift caller or the joint caller?

@arpanda
Copy link
Member

arpanda commented Feb 5, 2024

Yes, the default is not set for the bimodal hypothesis as you can see the false is hard coded in the code.

The mean shift caller has the limitation that its based on read depth only. For the supplementary figure S1, the normalization is based on joint caller.

-Arijit

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants