Closed elimereu closed 5 years ago
Hi,
actually I solved it. The problem was related to the fact that the count matrices wasn't a matrix object but a sparse matrix. And for the second problem.. I was using a factor instead to use a character vector for the seg_mouse. Now, it works fine! Thanks a lot!
Elisabetta
2018-09-23 11:13 GMT+02:00 kevinwang notifications@github.com:
Hi,
Thanks for letting us know. My colleague is looking at the first issue of union for you. I had a look at the second issue, I am suspecting that there is a missing value either in
ctl = seg_mouse
orsce
, do you mind checking that? If not, could you also try to setfast_svd = FALSE
?Thank you Best Wishes Kevin
On 21 Sep 2018, at 6:30 pm, Elisabetta <notifications@github.com<mailto: notifications@github.com>> wrote:
Hi,
I'm testing your tool! The idea is really nice and I would like to use it in my analysis. Before apply it with my sample, I took 2 different tissues from the tabula muris sample (colon and marrow analysed with Smart-seq2) and I merged them in a unique sce object. For the combination of SCE object I wanted to use the method = "union", but it wasn't possible because of this error:
Error in mat[rownames(x), ] <- assay(x, exprs[i]) : number of items to replace is not a multiple of replacement length .
My idea of union was to include also genes that are present in one dataset but not in the other. Why is this happen?
So, I used the intersection way.
Then, with the combined SCE, I run this command:
norm=scMerge(sce, ctl = seg_mouse, kmeansK = c(10,10), exprs = "logcounts",fast_svd = TRUE,return_all_RUV = FALSE,assay_name = "norm")
The tool run for some minutes and I get this output and error:
_``` No maker nor marker_list information was supplied Finding HVG... [1] 3707 Clustering within each batch... Performing pca... Creating Mutual Nearest Cluster... [1] 1 group batch cluster 1 1 1 5 2 2 2 9 Dimension of the replicates mapping matrix [1] 8356 4203
Performing RUV normalisation... This might take a few minutes... Error in if (k > n) k <- n : missing value where TRUE/FALSE needed
Any helps?
Thank you,
Elisabetta
— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHubhttps://protect-au. mimecast.com/s/cgAECVAGXPtv32XMsGIG_F?domain=github.com, or mute the threadhttps://protect-au.mimecast.com/s/Z0C_CWLJY7ikO6x0sxEAfT?domain= github.com.
— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/SydneyBioX/scMerge/issues/1#issuecomment-423802553, or mute the thread https://github.com/notifications/unsubscribe-auth/ATrry_F-f4bysq6wWq0iKGeeK7w-m0nUks5ud1DCgaJpZM4Wzr1e .
Hi,
Thanks for letting us know. My colleague is looking at the first issue of union for you. I had a look at the second issue, I am suspecting that there is a missing value either in
ctl = seg_mouse
orsce
, do you mind checking that? If not, could you also try to setfast_svd = FALSE
?Thank you Best Wishes Kevin
On 21 Sep 2018, at 6:30 pm, Elisabetta notifications@github.com<mailto:notifications@github.com> wrote:
Hi,
I'm testing your tool! The idea is really nice and I would like to use it in my analysis. Before apply it with my sample, I took 2 different tissues from the tabula muris sample (colon and marrow analysed with Smart-seq2) and I merged them in a unique sce object. For the combination of SCE object I wanted to use the method = "union", but it wasn't possible because of this error:
My idea of union was to include also genes that are present in one dataset but not in the other. Why is this happen?
So, I used the intersection way.
Then, with the combined SCE, I run this command:
norm=scMerge(sce, ctl = seg_mouse, kmeansK = c(10,10), exprs = "logcounts",fast_svd = TRUE,return_all_RUV = FALSE,assay_name = "norm")
The tool run for some minutes and I get this output and error:
_``` No maker nor marker_list information was supplied Finding HVG... [1] 3707 Clustering within each batch... Performing pca... Creating Mutual Nearest Cluster... [1] 1 group batch cluster 1 1 1 5 2 2 2 9 Dimension of the replicates mapping matrix [1] 8356 4203
Performing RUV normalisation... This might take a few minutes... Error in if (k > n) k <- n : missing value where TRUE/FALSE needed
Any helps?
Thank you,
Elisabetta
— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHubhttps://protect-au.mimecast.com/s/cgAECVAGXPtv32XMsGIG_F?domain=github.com, or mute the threadhttps://protect-au.mimecast.com/s/Z0C_CWLJY7ikO6x0sxEAfT?domain=github.com.