Support BiocSingular’s LowRankMatrix

[LuckyMD]

Hi!

I ran into a strange error when trying to write a quick script for Seurat -> Anndata conversion. I'm just loading data from a Seurat tutorial and then want to use their convert function to SingleCellExperiment, but anndata2ri is complaining. Here's the code:

%%R
library(Seurat)

# Load PBMC dataset from Seurat tutorial
pbmc.data <- Read10X(data.dir = "./filtered_gene_bc_matrices/hg19/")

# Initialize the Seurat object with the raw (non-normalized data).
pbmc <- CreateSeuratObject(counts = pbmc.data, project = "pbmc3k", min.cells = 3, min.features = 200)

%%R 
#convert the Seurat object to a SingleCellExperiment object
pbmc_sce <- as.SingleCellExperiment(pbmc)

pbmc_sce

In the second cell, anndata2ri complains, although I have not put in any -o flags to output the object to python. The error message is the following:

---------------------------------------------------------------------------
AttributeError                            Traceback (most recent call last)
<ipython-input-12-3fcf8ce65415> in <module>
----> 1 get_ipython().run_cell_magic('R', '', '#convert the Seurat object to a SingleCellExperiment object\npbmc_sce <- as.SingleCellExperiment(pbmc)\n\npbmc_sce\n')

/usr/local/lib/python3.7/site-packages/IPython/core/interactiveshell.py in run_cell_magic(self, magic_name, line, cell)
   2350             with self.builtin_trap:
   2351                 args = (magic_arg_s, cell)
-> 2352                 result = fn(*args, **kwargs)
   2353             return result
   2354 

</usr/local/lib/python3.7/site-packages/decorator.py:decorator-gen-806> in R(self, line, cell, local_ns)

/usr/local/lib/python3.7/site-packages/IPython/core/magic.py in <lambda>(f, *a, **k)
    185     # but it's overkill for just that one bit of state.
    186     def magic_deco(arg):
--> 187         call = lambda f, *a, **k: f(*a, **k)
    188 
    189         if callable(arg):

/usr/local/lib/python3.7/site-packages/rpy2/ipython/rmagic.py in R(self, line, cell, local_ns)
    725                     return_output = False
    726             else:
--> 727                 text_result, result, visible = self.eval(code)
    728                 text_output += text_result
    729                 if visible:

/usr/local/lib/python3.7/site-packages/rpy2/ipython/rmagic.py in eval(self, code)
    257             try:
    258                 # Need the newline in case the last line in code is a comment.
--> 259                 value, visible = ro.r("withVisible({%s\n})" % code)
    260             except (ri.embedded.RRuntimeError, ValueError) as exception:
    261                 # Otherwise next return seems to have copy of error.

/usr/local/lib/python3.7/site-packages/rpy2/robjects/vectors.py in __getitem__(self, i)
    260 
    261         if isinstance(res, Sexp):
--> 262             res = conversion.rpy2py(res)
    263         return res
    264 

/usr/local/Cellar/python/3.7.2_2/Frameworks/Python.framework/Versions/3.7/lib/python3.7/functools.py in wrapper(*args, **kw)
    822                             '1 positional argument')
    823 
--> 824         return dispatch(args[0].__class__)(*args, **kw)
    825 
    826     funcname = getattr(func, '__name__', 'singledispatch function')

/usr/local/lib/python3.7/site-packages/anndata2ri/r2py.py in rpy2py_s4(obj)
     25         return rpy2py_data_frame(obj)
     26     elif "SingleCellExperiment" in r_classes:
---> 27         return rpy2py_single_cell_experiment(obj)
     28     elif supported_r_matrix_classes() & r_classes:
     29         return rmat_to_spmat(obj)

/usr/local/lib/python3.7/site-packages/anndata2ri/r2py.py in rpy2py_single_cell_experiment(obj)
     56             assay_names = [str(a) for a in se.assayNames(obj)]
     57             # The assays can be stored in an env or elsewise so we don’t use obj.slots['assays']
---> 58             assays = [numpy2ri.rpy2py(assay).T for assay in (se.assay(obj, n) for n in assay_names)]
     59             # There’s SingleCellExperiment with no assays
     60             exprs, layers = assays[0], dict(zip(assay_names[1:], assays[1:]))

/usr/local/lib/python3.7/site-packages/anndata2ri/r2py.py in <listcomp>(.0)
     56             assay_names = [str(a) for a in se.assayNames(obj)]
     57             # The assays can be stored in an env or elsewise so we don’t use obj.slots['assays']
---> 58             assays = [numpy2ri.rpy2py(assay).T for assay in (se.assay(obj, n) for n in assay_names)]
     59             # There’s SingleCellExperiment with no assays
     60             exprs, layers = assays[0], dict(zip(assay_names[1:], assays[1:]))

AttributeError: 'RS4' object has no attribute 'T'

[LuckyMD]

thx, works!

[SabrinaRichter]

I get the same error message for a different problem, can't figure out what I am doing wrong, maybe someone could help me?
I am trying to get fastMNN running in my jupyter notebook, but it's giving me this error. The line before (logcounts(adata) ...) works fine.

import rpy2.robjects as robjects
from rpy2.robjects.packages import importr
import anndata2ri
anndata2ri.activate()
%load_ext rpy2.ipython

import anndata
import numpy as np
importr('batchelor')

X = np.random.random((10, 20))
adata = anndata.AnnData(X)
adata.obs['cohort'] = ['c1', 'c2'] * 5

%%R -i adata

logcounts(adata) <- assay(adata, 'X')
fastMNN(adata, batch=adata$cohort)

---------------------------------------------------------------------------
AttributeError                            Traceback (most recent call last)
<ipython-input-4-7c3b7739d5cc> in <module>
----> 1 get_ipython().run_cell_magic('R', '-i adata', "\nlogcounts(adata) <- assay(adata, 'X')\nfastMNN(adata, batch=adata$cohort)\n")

~/miniconda3/envs/clinspect/lib/python3.8/site-packages/IPython/core/interactiveshell.py in run_cell_magic(self, magic_name, line, cell)
   2379             with self.builtin_trap:
   2380                 args = (magic_arg_s, cell)
-> 2381                 result = fn(*args, **kwargs)
   2382             return result
   2383 

<decorator-gen-121> in R(self, line, cell, local_ns)

~/miniconda3/envs/clinspect/lib/python3.8/site-packages/IPython/core/magic.py in <lambda>(f, *a, **k)
    185     # but it's overkill for just that one bit of state.
    186     def magic_deco(arg):
--> 187         call = lambda f, *a, **k: f(*a, **k)
    188 
    189         if callable(arg):

~/miniconda3/envs/clinspect/lib/python3.8/site-packages/rpy2/ipython/rmagic.py in R(self, line, cell, local_ns)
    761                     return_output = False
    762             else:
--> 763                 text_result, result, visible = self.eval(code)
    764                 text_output += text_result
    765                 if visible:

~/miniconda3/envs/clinspect/lib/python3.8/site-packages/rpy2/ipython/rmagic.py in eval(self, code)
    266             try:
    267                 # Need the newline in case the last line in code is a comment.
--> 268                 value, visible = ro.r("withVisible({%s\n})" % code)
    269             except (ri.embedded.RRuntimeError, ValueError) as exception:
    270                 # Otherwise next return seems to have copy of error.

~/miniconda3/envs/clinspect/lib/python3.8/site-packages/rpy2/rinterface_lib/sexp.py in __iter__(self)
    662     def __iter__(self) -> typing.Iterator[typing.Union[Sexp, VT, typing.Any]]:
    663         for i in range(len(self)):
--> 664             yield self[i]
    665 
    666     def index(self, item: typing.Any) -> int:

~/miniconda3/envs/clinspect/lib/python3.8/site-packages/rpy2/robjects/vectors.py in __getitem__(self, i)
    266 
    267         if isinstance(res, Sexp):
--> 268             res = conversion.rpy2py(res)
    269         return res
    270 

~/miniconda3/envs/clinspect/lib/python3.8/functools.py in wrapper(*args, **kw)
    873                             '1 positional argument')
    874 
--> 875         return dispatch(args[0].__class__)(*args, **kw)
    876 
    877     funcname = getattr(func, '__name__', 'singledispatch function')

~/miniconda3/envs/clinspect/lib/python3.8/site-packages/anndata2ri/r2py.py in rpy2py_s4(obj)
     26         return rpy2py_data_frame(obj)
     27     elif "SingleCellExperiment" in r_classes:
---> 28         return rpy2py_single_cell_experiment(obj)
     29     elif supported_r_matrix_classes() & r_classes:
     30         return rmat_to_spmat(obj)

~/miniconda3/envs/clinspect/lib/python3.8/site-packages/anndata2ri/r2py.py in rpy2py_single_cell_experiment(obj)
     74             assay_names = [str(a) for a in se.assayNames(obj)]
     75             # The assays can be stored in an env or elsewise so we don’t use obj.slots['assays']
---> 76             assays = [mat_converter.rpy2py(assay).T for assay in (se.assay(obj, n) for n in assay_names)]
     77             # There’s SingleCellExperiment with no assays
     78             exprs, layers = assays[0], dict(zip(assay_names[1:], assays[1:]))

~/miniconda3/envs/clinspect/lib/python3.8/site-packages/anndata2ri/r2py.py in <listcomp>(.0)
     74             assay_names = [str(a) for a in se.assayNames(obj)]
     75             # The assays can be stored in an env or elsewise so we don’t use obj.slots['assays']
---> 76             assays = [mat_converter.rpy2py(assay).T for assay in (se.assay(obj, n) for n in assay_names)]
     77             # There’s SingleCellExperiment with no assays
     78             exprs, layers = assays[0], dict(zip(assay_names[1:], assays[1:]))

AttributeError: 'RS4' object has no attribute 'T'

[emdann]

Exact same problem here running fastMNN within Jupyter notebook. Things I tried that didn't work:

• saving the SingleCellExperiment object as RDS and reloading (works within R)
• Running library(SingleCellExperiment) before conversion
• converting all matrices in assays to dgRMatrix

[mrland99]

Was this ever resolved? Having same problem running fastMNN

[flying-sheep]

I added some code that helps with error messages in #82, now we can see

TypeError: Cannot convert assay “reconstructed” of type(s) [LowRankMatrix] to Python

If you don’t need it, you can delete that assay before converting.

Otherwise we need to add support for that class, any idea what would fit?

I assume it’s this class: https://rdrr.io/bioc/BiocSingular/man/LowRankMatrix.html

[LUMS-madiha]

Hi, if anyone here could manage to resolve that "AttributeError: 'RS4' object has no attribute 'T'" for fastMNN?