JunyueC
commented
5 years ago Thanks for your interest in the data.
All samples in MOCA are processed in a single sci-RNA-seq3 experiment and sequenced in one NOVA-seq run. Besides, embryos from different development stages are randomized for nuclei extraction. As a result, we do not see obvious batch effect in the data - this is further validated by the overlapping of cells from >10 individual replicates of the same development stage. Thus we did not include batch correction in the processing pipeline as this may remove real biological variances..
Jun
On Jun 15, 2019, at 10:40 PM, baibai notifications@github.com wrote:
Hi, In your paper (The single-cell transcriptional landscape of mammalian organogenesis https://www.nature.com/articles/s41586-019-0969-x#Abs1),I'm wondering how you make mouse embryos data batch corrected among different times. Batch effect remove is very important in sc-RNAseq. You just referred one time in your article that Mouse embryos from different development stages were processed together to reduce batch effects https://www.nature.com/articles/s41586-019-0969-x#Sec8.But these samples can't be sequenced in a lane.So they still contain batch difference remain to be corrected.So can you share your method to treat batch effect?Why didn't you use Monole or Seurat to deal with it?And I saw the parameter (residual_model_formula_str = "~ batch") in monocle can hold it.But Seurat's batch corrected result ([Comprehensive Integration of Single Cell Dat](Comprehensive Integration of Single Cell Data:)a )seems much better than monocle does. I'm really expecting your reply!
— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/JunyueC/sci-RNA-seq3_pipeline/issues/1?email_source=notifications&email_token=ADSB5ODNUOM5ZGVOXFNNFBLP2XG4FA5CNFSM4HYQNQM2YY3PNVWWK3TUL52HS4DFUVEXG43VMWVGG33NNVSW45C7NFSM4GZXHSRQ, or mute the thread https://github.com/notifications/unsubscribe-auth/ADSB5OHBDIBDFNKASGZUJQTP2XG4FANCNFSM4HYQNQMQ.
Hi, In your paper (The single-cell transcriptional landscape of mammalian organogenesis),I'm wondering how you make mouse embryos data batch corrected among different times. Batch effect remove is very important in sc-RNAseq. You just referred one time in your article that Mouse embryos from different development stages were processed together to reduce batch effects.But these samples can't be sequenced in a lane.So they still contain batch difference remain to be corrected.So can you share your method to treat batch effect?Why didn't you use Monole or Seurat to deal with it?And I saw the parameter (residual_model_formula_str = "~ batch") in monocle can hold it.But Seurat's batch corrected result (Comprehensive Integration of Single Cell Data)seems much better than monocle does after my attempt. I'm really expecting your reply!