Open bio-xtt opened 4 days ago
Hi @bio-xtt ,
Can you check the 'Unique_BATCH_train_5_80.npz' file? It looks like this file is empty. And not only this file, it looks like all files with size= 4kb are empty.
Would you be able to run our sample data using this sample script?
I would suggest you to first check the directory and parameters used to build coverage vector are properly used.
Thanks, Yumin
hi @bio-xtt ,
This is the output of our sample pipeline.
4.0K 5_80_stat.csv 12K M&U_5_80%.pkl 4.0K Multi_Batch_train_5_80.npz 360K Multi_meta_train_5_80.pkl 32K Multi_Region_train_5_80.npz 144K Multi_TE_train_5_80.npz 4.0K Unique_BATCH_train_5_80.npz 24K Unique_selected_meta_5_80.pkl 96K Unique_TE_train_5_80.npz
You can check your input files, parameters, and directory using our example data and pipeline.
Thanks, Yumin
Is this due to my improper setting of parameters? How should I adjust my parameters in this case?
Hi @bio-xtt ,
If you may, can you share a part of your aligned data with us? Including cell barcode (like the test_cb.tsv file), bam, and bam.bai file.
Thanks, Yumin
Thank you very much. I captured the cellrannger bam file for each sample chr1:1-1000000 and provided the barcodes.tsv file in the filtered_feature_bc_matrix folder. data download link: https://zenodo.org/api/records/13969187/draft/files/test_data.zip/content
Hi,
Please grant the permission for me to access this data.
Thanks, Yumin
Can you try this link below? https://gofile.io/d/WxRkct
Hello, my previous code is running normally, but the following error occurs when I run MATES_model.train. How should I solve it? code
MATES_model.train('10X', 'human_sample_list_file.txt', bin_size = 5, proportion = 80, BATCH_SIZE= 4096, AE_LR = 1e-4, MLP_LR = 1e-6, AE_EPOCHS = 200, MLP_EPOCHS = 200, DEVICE='cpu')
error
directory: