1
$\begingroup$

I am trying to combine two datasets to be able to compare the cells in them and using Seurat CCA for it:

https://satijalab.org/seurat/Seurat_AlignmentTutorial.html

When I am running AlignSubspace function I am getting a warning for each single dimension that I selected it to be run on:

seurat_union <- AlignSubspace(object = seurat_union, reduction.type = 
    "cca", grouping.var = "protocol", dims.align = 1:13) 

So, I am getting 13 warnings:

> warnings()
Warning messages:
1: In dtw(x = align.1, y = align.2, keep.internals = TRUE,  ... :
Argument dist.method is only useful with multivariate timeseries
2: In dtw(x = align.1, y = align.2, keep.internals = TRUE,  ... :
Argument dist.method is only useful with multivariate timeseries
3: In dtw(x = align.1, y = align.2, keep.internals = TRUE,  ... :
Argument dist.method is only useful with multivariate timeseries
4: In dtw(x = align.1, y = align.2, keep.internals = TRUE,  ... :
Argument dist.method is only useful with multivariate timeseries
5: In dtw(x = align.1, y = align.2, keep.internals = TRUE,  ... :
Argument dist.method is only useful with multivariate timeseries
6: In dtw(x = align.1, y = align.2, keep.internals = TRUE,  ... :
Argument dist.method is only useful with multivariate timeseries
7: In dtw(x = align.1, y = align.2, keep.internals = TRUE,  ... :
Argument dist.method is only useful with multivariate timeseries
8: In dtw(x = align.1, y = align.2, keep.internals = TRUE,  ... :
Argument dist.method is only useful with multivariate timeseries
9: In dtw(x = align.1, y = align.2, keep.internals = TRUE,  ... :
Argument dist.method is only useful with multivariate timeseries
10: In dtw(x = align.1, y = align.2, keep.internals = TRUE,  ... :
Argument dist.method is only useful with multivariate timeseries
11: In dtw(x = align.1, y = align.2, keep.internals = TRUE,  ... :
Argument dist.method is only useful with multivariate timeseries
12: In dtw(x = align.1, y = align.2, keep.internals = TRUE,  ... :
Argument dist.method is only useful with multivariate timeseries
13: In dtw(x = align.1, y = align.2, keep.internals = TRUE,  ... :
Argument dist.method is only useful with multivariate timeseries

What is this warning telling me? What should I pay attention to with my datasets?

$\endgroup$
5
  • $\begingroup$ Could you please clarify if you had these warnings running the tutorial or with your own dataset? If it is the second case could you please provide information about your data if it is a time series or not, and the type of data you add as an input? $\endgroup$
    – llrs
    Jun 8, 2018 at 11:02
  • $\begingroup$ With my own dataset. The input is two previously saved Seurat objects. I did analysis of the datasets separately, but then needed to combine them to compare. The input to Seurat objects were raw expression matrices for single cell rna seq data. $\endgroup$ Jun 8, 2018 at 16:45
  • $\begingroup$ And are you sure that seurat_union is the type of the object it is needed. It mind if it wouldn't be you would get an error, but a warning might indicate that perhaps you need to add some other type of data with this object or with other attributes. But it simply might be that the dims.align parameter should be something else $\endgroup$
    – llrs
    Jun 8, 2018 at 17:24
  • $\begingroup$ I am following the tutorial step by step except that I have already treated the input Seurat objects: normalized them, did scaling, clustering, tsne, etc. they way it is specified in a different Seurat tutorial. $\endgroup$ Jun 8, 2018 at 18:22
  • $\begingroup$ I am able to align them, and work further, it does not give any other error down the line, but I am afraid that something still went wrong, the alignment was not as good as expected and maybe I will get some result at the end but it may be flawed because I did not take into account this warning - that is what I am afraid of $\endgroup$ Jun 8, 2018 at 18:23

1 Answer 1

1
$\begingroup$

I don't have enough reputation to comment so I will briefly answer here

This warning is fixed in Seurat 2.3.2. Without upgrading you can ignore this message. It doesn't mean anything for the actual analysis performed so your results won't be flawed.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.