error on the transfer learning step

[helianthuszhu]

Hi Junyi,

 Lucky to see the scDEAL pipeline which is potentially useful for my drug associated scRNA data.
 Here, I am trying the pipeline on my dataset. The bulkmode.py step can be successfully finished with the output in the 'save/bulk_pre/' or 'save/bulk_encoder/' and ROC.pdf. However, when it comes to the scmodel.py step, something goes wrong as below. Any suggestions on this would be appreciated. Thank you very much.

step1 (can finish)

python bulkmodel.py --data './data/tnfa_batched_after/expM_batched.csv' \ --label './data/tnfa_batched_after/clin_batched.csv' \ --dataname 'GSEbatch' \ --result './save/results/result' --drug 'TNFA' \ --dimreduce 'AE' --encoder_h_dims "256,256" --predictor_h_dims "128,64" \ --bottleneck 256 --sampling 'SMOTE' --dropout 0.3 --printgene 'F' -mod 'new' #

step2

python scmodel.py --bulk_data './data/expM_batched.csv' \ --label './data/clin_batched.csv' \ --sc_data '/data/adata_rawcount_matrix.csv' \ --bulk_model_path 'save/bulk_pre' \ --pretrain 'save/sc_encoder' \ --dimreduce 'AE' --drug 'TNFA' \ --percent_mito 10 \ --bulk_h_dims "256,256" --bottleneck 256 --predictor_h_dims "128,64" --dropout 0.3 --printgene 'F' -mod 'new'

error

Traceback (most recent call last): File "scmodel.py", line 794, in run_main(args) File "scmodel.py", line 302, in run_main para = args.bulk+"data"+args.sc_data+"drug"+args.drug+"bottle"+str(args.bottleneck)+"edim"+args.bulk_h_dims+"pdim"+args.predictor_h_dims+"model"+reduce_model+"dropout"+str(args.dropout)+"gene"+args.printgene+"lr"+str(args.lr)+"mod"+args.mod+"sam"+args.sampling TypeError: can only concatenate str (not "NoneType") to str

I also tried to input the scRNA matrix from your example, like GSE112274_cell_gene_FPKM.csv, but the same error occured.

[FLabbaf97]

Hello, were you able to solve this problem?

[gwyang6]

Hello, thank you very much for your project. I have encountered the same problem. Can you solve it?

[juychen]

Hi, I have updated the code accordingly

Hello, thank you very much for your project. I have encountered the same problem. Can you solve it?

Hello, were you able to solve this problem?

Hi Junyi,

 Lucky to see the scDEAL pipeline which is potentially useful for my drug associated scRNA data.
 Here, I am trying the pipeline on my dataset. The bulkmode.py step can be successfully finished with the output in the 'save/bulk_pre/' or 'save/bulk_encoder/' and ROC.pdf. However, when it comes to the scmodel.py step, something goes wrong as below. Any suggestions on this would be appreciated. Thank you very much.

step1 (can finish) python bulkmodel.py --data './data/tnfa_batched_after/expM_batched.csv' --label './data/tnfa_batched_after/clin_batched.csv' --dataname 'GSEbatch' --result './save/results/result' --drug 'TNFA' --dimreduce 'AE' --encoder_h_dims "256,256" --predictor_h_dims "128,64" --bottleneck 256 --sampling 'SMOTE' --dropout 0.3 --printgene 'F' -mod 'new'

step2 python scmodel.py --bulk_data './data/expM_batched.csv' --label './data/clin_batched.csv' --sc_data '/data/adata_rawcount_matrix.csv' --bulk_model_path 'save/bulk_pre' --pretrain 'save/sc_encoder' --dimreduce 'AE' --drug 'TNFA' --percent_mito 10 --bulk_h_dims "256,256" --bottleneck 256 --predictor_h_dims "128,64" --dropout 0.3 --printgene 'F' -mod 'new' #error Traceback (most recent call last): File "scmodel.py", line 794, in run_main(args) File "scmodel.py", line 302, in run_main para = args.bulk+"data"+args.sc_data+"drug"+args.drug+"bottle"+str(args.bottleneck)+"edim"+args.bulk_h_dims+"pdim"+args.predictor_h_dims+"model"+reduce_model+"dropout"+str(args.dropout)+"gene"+args.printgene+"lr"+str(args.lr)+"mod"+args.mod+"sam"+args.sampling TypeError: can only concatenate str (not "NoneType") to str

I also tried to input the scRNA matrix from your example, like GSE112274_cell_gene_FPKM.csv, but the same error occured.