Comparing more than two datasets about liger HOT 12 CLOSED

Currently, the only way to do what you're asking is to create a liger object with 2 datasets--healthy and tumor--in which the tumor dataset contains the stacked DGE matrices from the 3 replicates. We are thinking about other ways to plot and interpret dataset-specific metagenes in cases with more than 2 groups, but don't have anything in the package yet.

from liger.

Thank you very much! I will try what you have suggested.

from liger.

@jw156605
Hi, are there any updates on this issue?

For example, is it possible to integrate datesets of 4 covid patients and 4 healthy controls?

from liger.

For that comparison, you could try two things: (1) combine the 4 covid patients into a single dataset (learn one H matrix for all 4) and do the same for the healthy controls or (2) treat each patient as a separate dataset and integrate all 8. We used option 2 in the Cell paper with data from 7 human brains.

from liger.

@jw156605
Thank you for your answer.

from liger.

@jw156605

Dear Welch,

Is the R code available for the Cell paper? I am a student and relatively new in this field. Learning from the R script file would be so much easy for me.

from liger.

@robin1-2 You can just paste an unique prefix before all the barcodes in each of the 8 datasets. For example, you can paste "covid1-" to the first patient dataset and "ctrl1-" to the first healty control. Then you can treat them as 8 different datasets and follow the tutorials we provided.

from liger.

@skpalan
Thank you so much for your answer. Two tutorial shows the integration of two datasets, I get confused when I tried to intergrate more.
Is the following code true?

ifnb_liger <- createLiger(list(ctrl = list(ctrl1, ctrl2, ctrl3, ctrl4), stim = list(stim1, stim2, stim3, stim4))

from liger.

@robin1-2 I would recommend:

# "ctrl1" --> data name passed to LIGER, can change to any names you prefer
# "control1" --> your dataset as matrix
data.list <- list(ctrl1 = control1, ctrl2 = control2, ctrl3 = control3, ctrl4 = control4, cov1 = covid1, cov2 = covid2, cov3 = covid3, cov4 = covid4)
ifnb_liger <- createLiger(data.list)

from liger.

@skpalan
I see! Thank you so much.

from liger.

@jw156605
dear Welch, i am very confused about the downstream analysis after creating the liger object which consist of 2 ctrls and 2 covid patients. Is the r script of your Cell paper with brain samples available?

from liger.

from liger.