search

broadinstitute / infercnv

Inferring CNV from Single-Cell RNA-Seq
Other
557 stars 164 forks source link

error when using HMM=T #534

Open mayan2009 opened 1 year ago

mayan2009 commented 1 year ago

Hello

I am runing infercnv for a dataset with 50K cells and I used the parameter HMM=T, Then I got stuck in this procedure. INFO [2023-05-03 02:15:56] Gibbs sampling time: 451.820352248351 Minutes

It is already for almost 30 hours. but it is still here. Could anyone tell me does this step do need such a long time or maybe there is something wrong with my code. I already tried the same code and dataset but with HMM=F and it worked well.

GeorgescuC commented 1 year ago

Hi @mayan2009 ,

The Bayesian network step can take a while to run depending on how many CNV regions have been predicted by the HMM. The issue might be that you have too many small subclusters resulting in too many CNV regions. If you check the earlier log lines, how many CNVs are predicted?

Regards, Christophe.