search

edroaldo / cellrouter

Reconstruction of complex single-cell trajectories using CellRouter
45 stars 21 forks source link

suitable Normalization for cellrouter - scran vs scone vs SCnorm #17

Closed AswinSSoman closed 2 years ago

edroaldo commented 5 years ago

Please, check my previous answer. In several cases, I updated the slot cellrouter@ndata to use normalized data from other sources.

Hope it helps!

Thanks!

Em qua, 21 de nov de 2018 às 02:50, Aswin S Soman notifications@github.com escreveu:

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/edroaldo/cellrouter/issues/17, or mute the thread https://github.com/notifications/unsubscribe-auth/AJqUR3h-eJIcp6oofln9X9jlJiLrQcpQks5uxQXCgaJpZM4Ysk1n .

-- Edroaldo

AswinSSoman commented 5 years ago

Hi edroaldo,

Thank you for your reply.

Is it possible to add multiple normalised matrices of same expression data in cellrouter@ndata without overwriting the existing matrices, so that I can compare how normalised data affected the results within cellrouter?

sorry for the delayed response.

Best regards, Aswin

edroaldo commented 5 years ago

This is not possible in the current version of CellRouter. I am planning a release in a few weeks (but maybe sooner) where I could include this feature.

Thanks!

Em qui, 29 de nov de 2018 às 03:07, Aswin S Soman notifications@github.com escreveu:

Hi edroaldo,

Thank you for your reply.

Is it possible to add multiple normalised matrices of same expression data in cellrouter@ndata without overwriting the existing matrices, so that I can compare how normalised data affected the results within cellrouter?

sorry for the delayed response.

Best regards, Aswin

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/edroaldo/cellrouter/issues/17#issuecomment-442742304, or mute the thread https://github.com/notifications/unsubscribe-auth/AJqURyQYA5RfW2ADkvMsZSPU20IOHPHGks5uz5XXgaJpZM4Ysk1n .

-- Edroaldo

AswinSSoman commented 5 years ago

Is it possible to do external clustering and add data to 'graph' & 'sampTab' slots in cellrouter object?

edroaldo commented 5 years ago

I am sorry for my delayed response. Yes, you can do this using the addInfo function. Assuming that metadata is a daframe that contais a column "cluster" with your clusters, you can do: cellrouter <- addInfo(cellrouter, metadata = metadata, colname = 'cluster', metadata.column = 'cluster) where colname='cluster' will add a column named ''cluster' to cellrouter@sampTab. Then, you can easily using this metadata for any other downstream analysis, for example as long as you specify the column to use using the parameter "annotation" and annotation.color:

plotReducedDimension(cellrouter, reduction.type = 'tsne', dims.use = c(1,2), annotation = "cluster", annotation.color = 'cluster_color', showlabels = TRUE, 4.5, 3.5, filename='results/tSNE_2.pdf')

The tutorial https://github.com/edroaldo/cellrouter/blob/master/Myeloid_Progenitors/CellRouter_Paul_Tutorial.md illustrates how to use different metadata to create different Knn representations, for example: cellrouter <- buildKNN(cellrouter, k = 10, column.ann = 'cluster', num.pcs = 20, sim.type = 'jaccard') plotKNN(cellrouter, reduction.type = 'tsne', column.ann = 'cluster, column.color = 'cluster_color', width = 5, height = 4, filename='results/knn_tsne_original_clusters.pdf')

Thanks!

Em sex, 30 de nov de 2018 às 00:43, Aswin S Soman notifications@github.com escreveu:

Is it possible to do external clustering and add data to 'graph' & 'sampTab' slots in cellrouter object?

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/edroaldo/cellrouter/issues/17#issuecomment-443097896, or mute the thread https://github.com/notifications/unsubscribe-auth/AJqURxEFQQ7dXhAJASC691PBrz-DcFGRks5u0MV2gaJpZM4Ysk1n .

-- Edroaldo