A bug in rctd4weight

[EddieLv]

The full mode function is not suitable for my RCTD object!
My RCTD(full mode for multiple replicates!) object structure is like this:
Formal class 'RCTD.replicates' [package "spacexr"] with 5 slots
..@ RCTD.reps :List of 2
.. ..$ E12.5_slice13:Formal class 'RCTD' [package "spacexr"] with 9 slots
.. .. .. ..@ spatialRNA :Formal class 'SpatialRNA' [package "spacexr"] with 3 slots
.. .. .. ..@ originalSpatialRNA:Formal class 'SpatialRNA' [package "spacexr"] with 3 slots
.. .. .. ..@ reference :Formal class 'Reference' [package "spacexr"] with 3 slots
.. .. .. ..@ config :List of 22
.. .. .. .. ..$ gene_cutoff : num 0.000125
.. .. .. .. ..$ fc_cutoff : num 0.5
.. .. .. .. ..$ gene_cutoff_reg : num 2e-04
.. .. .. .. ..$ fc_cutoff_reg : num 0.75
.. .. .. .. ..$ UMI_min : num 100
.. .. .. .. ..$ UMI_min_sigma : num 300
.. .. .. .. ..$ max_cores : num 8
.. .. .. .. ..$ N_epoch : num 8
.. .. .. .. ..$ N_X : num 50000
.. .. .. .. ..$ K_val : num 100
.. .. .. .. ..$ N_fit : num 1000
.. .. .. .. ..$ N_epoch_bulk : num 30
.. .. .. .. ..$ MIN_CHANGE_BULK : num 1e-04
.. .. .. .. ..$ MIN_CHANGE_REG : num 0.001
.. .. .. .. ..$ UMI_max : num 2e+07
.. .. .. .. ..$ counts_MIN : num 10
.. .. .. .. ..$ MIN_OBS : num 3
.. .. .. .. ..$ MAX_MULTI_TYPES : num 4
.. .. .. .. ..$ CONFIDENCE_THRESHOLD: num 10
.. .. .. .. ..$ DOUBLET_THRESHOLD : num 25
.. .. .. .. ..$ RCTDmode : chr "full"
.. .. .. .. ..$ doublet_mode : chr "full"
.. .. .. ..@ cell_type_info :List of 2
.. .. .. .. ..$ info :List of 3
.. .. .. .. ..$ renorm:List of 3
.. .. .. ..@ internal_vars :List of 8
.. .. .. .. ..$ gene_list_reg : chr [1:1113] "Sulf1" "Col3a1" "Col5a2" "Fn1" ...
.. .. .. .. ..$ gene_list_bulk : chr [1:2640] "Sulf1" "Gulp1" "Col3a1" "Col5a2" ...
.. .. .. .. ..$ proportions : Named num [1:13] 0.12924 0.00166 0.0278 0.06929 0.0024 ...
.. .. .. .. .. ..- attr(, "names")= chr [1:13] "adven.FB" "art.EC" "Cardiomyocyte" "gCap" ...
.. .. .. .. ..$ class_df :'data.frame': 13 obs. of 1 variable:
.. .. .. .. ..$ cell_types_assigned: logi TRUE
.. .. .. .. ..$ sigma : num 0.56
.. .. .. .. ..$ Q_mat : num [1:103, 1:2536] 1.00 2.46e-05 4.96e-06 2.99e-06 2.11e-06 ...
.. .. .. .. .. ..- attr(, "dimnames")=List of 2
.. .. .. .. ..$ X_vals : num [1:2536] 1.00e-05 2.83e-05 5.20e-05 8.00e-05 1.12e-04 ...
.. .. .. ..@ results :List of 1
.. .. .. .. ..$ weights:Formal class 'dgCMatrix' [package "Matrix"] with 6 slots
.. .. .. ..@ de_results : list()
.. .. .. ..@ internal_vars_de : list()
.. ..$ E12.5_slice21:Formal class 'RCTD' [package "spacexr"] with 9 slots
.. .. .. ..@ spatialRNA :Formal class 'SpatialRNA' [package "spacexr"] with 3 slots
.. .. .. ..@ originalSpatialRNA:Formal class 'SpatialRNA' [package "spacexr"] with 3 slots
.. .. .. ..@ reference :Formal class 'Reference' [package "spacexr"] with 3 slots
.. .. .. ..@ config :List of 22
.. .. .. .. ..$ gene_cutoff : num 0.000125
.. .. .. .. ..$ fc_cutoff : num 0.5
.. .. .. .. ..$ gene_cutoff_reg : num 2e-04
.. .. .. .. ..$ fc_cutoff_reg : num 0.75
.. .. .. .. ..$ UMI_min : num 100
.. .. .. .. ..$ UMI_min_sigma : num 300
.. .. .. .. ..$ max_cores : num 8
.. .. .. .. ..$ N_epoch : num 8
.. .. .. .. ..$ N_X : num 50000
.. .. .. .. ..$ K_val : num 100
.. .. .. .. ..$ N_fit : num 1000
.. .. .. .. ..$ N_epoch_bulk : num 30
.. .. .. .. ..$ MIN_CHANGE_BULK : num 1e-04
.. .. .. .. ..$ MIN_CHANGE_REG : num 0.001
.. .. .. .. ..$ UMI_max : num 2e+07
.. .. .. .. ..$ counts_MIN : num 10
.. .. .. .. ..$ MIN_OBS : num 3
.. .. .. .. ..$ MAX_MULTI_TYPES : num 4
.. .. .. .. ..$ CONFIDENCE_THRESHOLD: num 10
.. .. .. .. ..$ DOUBLET_THRESHOLD : num 25
.. .. .. .. ..$ RCTDmode : chr "full"
.. .. .. .. ..$ doublet_mode : chr "full"
.. .. .. ..@ cell_type_info :List of 2
.. .. .. .. ..$ info :List of 3
.. .. .. .. ..$ renorm:List of 3
.. .. .. ..@ internal_vars :List of 8
.. .. .. .. ..$ gene_list_reg : chr [1:1112] "Sulf1" "Col3a1" "Col5a2" "Fn1" ...
.. .. .. .. ..$ gene_list_bulk : chr [1:2634] "Sulf1" "Gulp1" "Col3a1" "Col5a2" ...
.. .. .. .. ..$ proportions : Named num [1:13] 0.17447 0.01783 0.02557 0.05497 0.00442 ...
.. .. .. .. .. ..- attr(, "names")= chr [1:13] "adven.FB" "art.EC" "Cardiomyocyte" "gCap" ...
.. .. .. .. ..$ class_df :'data.frame': 13 obs. of 1 variable:
.. .. .. .. ..$ cell_types_assigned: logi TRUE
.. .. .. .. ..$ sigma : num 0.86
.. .. .. .. ..$ Q_mat : num [1:103, 1:2536] 1.00 3.78e-05 8.76e-06 5.24e-06 3.67e-06 ...
.. .. .. .. .. ..- attr(, "dimnames")=List of 2
.. .. .. .. ..$ X_vals : num [1:2536] 1.00e-05 2.83e-05 5.20e-05 8.00e-05 1.12e-04 ...
.. .. .. ..@ results :List of 1
.. .. .. .. ..$ weights:Formal class 'dgCMatrix' [package "Matrix"] with 6 slots
.. .. .. ..@ de_results : list()
.. .. .. ..@ internal_vars_de : list()
..@ population_de_results : list()
..@ population_sig_gene_list: list()
..@ population_sig_gene_df : list()
..@ group_ids : Named num [1:2] 1 1
.. ..- attr(*, "names")= chr [1:2] "E12.5_slice13" "E12.5_slice21"

[Vanjia-lee]

THX for your feedback. Here, the integration of 'full mode' results in the rctd4weight function is designed for a single RCTD object. For a list of RCTD objects, you may need to make changes to the function according to your data structure.