Input data production trouble

[Maryne972]

Hello,

Thank you for your great tools. I have a question: I'm not quite sure how to generate the input matrix from the normalized count table for RNA-Seq, ATAC-Seq, and WBGS data. Could you please explain?
Following an exemple of my data :
For ATAC-seq:

"Chr","Start","End","Bd21-3 D0r1","Bd21-3 D1r1","Bd21-3 D3r1","Bd21-3 D4r1","Bd21-3 D0r2","Bd21-3 D1r2","Bd21-3 D3r2","Bd21-3 D4r2","Bd21-3 D0r3","Bd21-3 D1r3","Bd21-3 D3r3","Bd21-3 D4r3","NA","NA","NA"
1,36222,36622,1,36222,36622,3927.05764014892,3436.04536976362,1978.58390924122,3424.18258835817,4982.43569281856,3728.81215775087,2863.17505667422,4117.86170355401,5108.41645644849,2436.39845482816,1780.36887556699,1716.3170845623
1,61084,61484,1,61084,61484,208.586179565762,129.007399952285,129.64306232078,129.103713699879,139.610953060372,154.586433675601,119.073442810648,103.639480979118,169.635402721379,122.136573116154,131.328532226932,141.551924499984
1,82192,82592,1,82192,82592,152.211536439881,124.558868919448,117.425391526151,132.03788901124,141.844728309338,182.693057980256,108.248584373316,79.5372761002531,146.8655500071,148.373318452216,117.700099637345,111.472140543737
1,82947,83347,1,82947,83347,279.618229904373,161.036823388714,191.410175782512,136.439151978282,254.650378382119,229.537431821348,142.269568033502,126.536575614039,260.714813578495,173.705348431863,155.488026363018,111.472140543737
1,110714,111114,1,110714,111114,515.264238170559,457.308990175686,359.063769464358,374.107352198514,529.404734004931,496.550362715568,398.97335383308,350.68708098748,555.584406228408,446.92938606948,402.658235601442,318.491830124963
1,165233,165633,1,165233,165633,214.22364387835,199.294190271116,142.539492603999,127.636626044199,193.221559035555,159.270871059711,114.434217766077,110.870142442777,198.097718614227,196.323232342262,134.42590327002,161.900013646856
1,215566,215966,1,215566,215966,417.172359131525,426.169272945824,281.006428276454,225.931498974789,489.196779523544,435.65267672215,298.456811200715,333.815537572274,391.641466685599,373.647442199788,282.480239129627,206.134990053101
1,230442,230842,1,230442,230842,153.339029302398,159.257410975579,116.746632037561,108.564486520353,151.896716929685,201.430807516693,112.114605243792,84.357717076026,125.234189928535,132.088442036729,90.44323445817,99.0863471499886
1,231432,231832,1,231432,231832,175.888886552751,104.095626168395,124.891745900646,148.175853223725,138.494065435889,168.639745827929,169.331714126831,257.893592203851,111.572278299967,108.565

RNA-seq and WGBS is the same with geneID instead peak ID.
thank you in advance

cheers

[jan-spr]

Hi,
to be able to help could you please clarify whether you are using mofapy2 in python, and not R? They both take lists of matrices, but the syntax is slightly different.
An example for python is here: https://github.com/bioFAM/mofapy2/blob/master/mofapy2/notebooks/getting_started_python.ipynb
In this case you need to pass the matrices as a nested list of np.arrays (see data_mat in the link), and the sample and feature names would have to be passed separately to ent.set_data_matrix(), e.g. as another nested list for features:

ent.set_data_matrix(
    data_mat,
    features_names = [feature_names_RNA,feature_names_ATAC,...]
)

For R there is info here: https://raw.githack.com/bioFAM/MOFA2_tutorials/master/R_tutorials/getting_started_R.html
hope this helps!

[Maryne972]

Hi,
Thanks you for your response. I'm using mofapy in python, I followed the example, but I don't understand what the value in the sample data represents.

sample   group   feature    value   view
sample1  groupA  gene1      2.8044  RNA
sample1  groupA  gene3      2.2069  RNA
sample2  groupB  gene2      0.1454  RNA
sample2  groupB  gene1      2.7021  RNA
sample2  groupB  promoter1  3.8618  Methylation
sample3  groupB  promoter2  3.2545  Methylation
sample3  groupB  promoter3  1.5014  Methylation

How do I enter my matrix list into the tool if the example file I showed you isn't the right one?

thank you in advance

cheers

[jan-spr]

Hi,
right, the notebook provides two input format examples. What you are looking at there is the data.frame format, part 2.2 instead shows a list of matrices example that is probably closer to your data formatting - though it is quite brief.

If you are still having trouble it could also be worth considering using mofa trough muon; if you create an AnnData object from each of your view matrices, you can combine them into a single MuData object, which then natively allows for running mofa trough mu.tl.mofa(). There's also a number of examples in the muon documentation, so hope you can make it work!