Error: `rownames(X)` is not as expected

[YinCY123]

sces_ann <- as_AnnData(x = sces, 
    x_mapping = "counts", 
    layers_mapping = c(logcounts = "logcounts"))

Error:
! rownames(X) is not as expected
ℹ Expected row names: "SRR8325947_AAACCTGTCAGGCAAG", "SRR8325947_AAACGGGAGCATCATC", "SRR8325947_AAACGGGCACCTTGTC", "SRR8325947_AAACGGGGTTTGTTTC",
"SRR8325947_AAACGGGTCGTTTATC", "SRR8325947_AAACGGGTCTGCGACG", "SRR8325947_AAAGATGAGCCAGAAC", "SRR8325947_AAAGATGAGGTGGGTT", "SRR8325947_AAAGATGCATATACGC",
"SRR8325947_AAAGCAAGTACCGCTG", "SRR8325947_AAAGCAATCCATGAAC", "SRR8325947_AAAGTAGAGAGGACGG", "SRR8325947_AAAGTAGAGCAGACTG", "SRR8325947_AAAGTAGCACATGGGA",
"SRR8325947_AAATGCCCATAGAAAC", "SRR8325947_AACCATGAGAGTAATC", "SRR8325947_AACCATGCAGCGAACA", "SRR8325947_AACCATGCAGTCAGCC", …, "SRR17008559_TTTGTTGTCTGCCTCA",
"SRR17008559_TTTGTTGTCTTTCGAT"
ℹ Provided row names: "1_AAACCTGTCAGGCAAG", "1_AAACGGGAGCATCATC", "1_AAACGGGCACCTTGTC", "1_AAACGGGGTTTGTTTC", "1_AAACGGGTCGTTTATC", "1_AAACGGGTCTGCGACG",
"1_AAAGATGAGCCAGAAC", "1_AAAGATGAGGTGGGTT", "1_AAAGATGCATATACGC", "1_AAAGCAAGTACCGCTG", "1_AAAGCAATCCATGAAC", "1_AAAGTAGAGAGGACGG", "1_AAAGTAGAGCAGACTG",
"1_AAAGTAGCACATGGGA", "1_AAATGCCCATAGAAAC", "1_AACCATGAGAGTAATC", "1_AACCATGCAGCGAACA", "1_AACCATGCAGTCAGCC", …, "130_TTTGTTGTCTGCCTCA", "130_TTTGTTGTCTTTCGAT"

I really don't know why arise this error, my barcode are all of this format SRR8325947_AAAGTAGAGCAGACTG, why this kind of barcode could appear 1_AAACGGGTCGTTTATC ?

And the sces is a SingleCellExperiment object

[lazappi]

Hi @YinCY123

It would be helpful to have some extra context on the object. Can you please check:

• What are the row/column names of the SingleCellExperiment?
• What are the row/column names of the "counts" matrix?
• What version of {anndataR} are you using?

If you can provide an example dataset that causes the error that would be the most helpful. It doesn't need to be the whole object but something like sces[1:10, 1:20] as long as you see the same error message.

[YinCY123]

@lazappi

I'm sorry for the delay.

1. the first 10 row names of my sces are:

 [1] "ENSG00000299200" "ENSG00000306402" "ENSG00000306474" "ENSG00000307061"
 [5] "ENSG00000307077" "ENSG00000300074" "ENSG00000300051" "ENSG00000307470"
 [9] "ENSG00000300846" "ENSG00000297432"

2. and the colum names are:

[1] "SRR8325947_AAACCTGTCAGGCAAG" "SRR8325947_AAACGGGAGCATCATC"
[3] "SRR8325947_AAACGGGCACCTTGTC" "SRR8325947_AAACGGGGTTTGTTTC"
[5] "SRR8325947_AAACGGGTCGTTTATC" "SRR8325947_AAACGGGTCTGCGACG"

3. the row and column names of counts are exactly same as the row and column names of sces
4. the version of anndataR i'm using is [1.0.0.]

The attached file is a subset of my sces object. As github do not allowed to attach qs format file, i changed the suffix, it can be read with qs::qread() in R.

Thanks for the great package you developed!

sces_small.gz

[lazappi]

@YinCY123

Thanks for providing the example dataset 🎉! I can confirm that something is causing the column names to be changed. Interestingly, it only happens when mapping to X.

This works:

> anndataR::as_AnnData(sce, x_mapping = NULL, layers_mapping = "counts")
InMemoryAnnData object with n_obs × n_vars = 10 × 10
    obs: 'batch', 'Sample', 'Barcode', 'samples', 'tissue', 'sum', 'detected', 'subsets_mito_sum', 'subsets_mito_detected', 'subsets_mito_percent', 'total', 'sizeFactor', 'tsne_nn_50', 'umap_nn_50', 'tsne_nn_100', 'umap_nn_100', 'tsne_nn_150', 'umap_nn_150', 'tsne_nn_200', 'umap_nn_200', 'celltype', 'subtype'
    var: 'rotation', 'ID', 'Symbol', 'Type', 'subsets_mito', 'mean', 'detected', 'NA', 'NA', 'NA', 'NA', 'NA', 'NA', 'NA', 'NA', 'NA', 'NA', 'NA', 'NA', 'NA', 'NA', 'NA', 'NA', 'NA', 'NA', 'NA', 'NA', 'NA', 'NA', 'NA', 'NA', 'NA', 'NA', 'NA', 'NA', 'NA', 'NA', 'NA', 'NA', 'NA', 'NA', 'NA', 'NA', 'NA', 'NA', 'NA', 'NA', 'NA', 'NA', 'NA', 'NA', 'NA', 'NA', 'NA', 'NA', 'NA'
    uns: 'merge.info', 'pca.info', 'Samples'
    obsm: 'corrected', 'TSNE', 'UMAP'
    layers: 'counts'

But this doesn't:

> anndataR::as_AnnData(sce, x_mapping = "counts", layers_mapping = FALSE)
Error:
! `rownames(X)` is not as expected
ℹ Expected row names: "SRR8325947_AAACCTGTCAGGCAAG", "SRR8325947_AAACGGGAGCATCATC", "SRR8325947_AAACGGGCACCTTGTC", "SRR8325947_AAACGGGGTTTGTTTC",
  "SRR8325947_AAACGGGTCGTTTATC", "SRR8325947_AAACGGGTCTGCGACG", "SRR8325947_AAAGATGAGCCAGAAC", "SRR8325947_AAAGATGAGGTGGGTT", "SRR8325947_AAAGATGCATATACGC",
  "SRR8325947_AAAGCAAGTACCGCTG"
ℹ Provided row names: "1_AAACCTGTCAGGCAAG", "1_AAACGGGAGCATCATC", "1_AAACGGGCACCTTGTC", "1_AAACGGGGTTTGTTTC", "1_AAACGGGTCGTTTATC", "1_AAACGGGTCTGCGACG",
  "1_AAAGATGAGCCAGAAC", "1_AAAGATGAGGTGGGTT", "1_AAAGATGCATATACGC", "1_AAAGCAAGTACCGCTG"
Run `rlang::last_trace()` to see where the error occurred.

We need to look more into what is happening here.