search

bowang-lab / scGPT

https://scgpt.readthedocs.io/en/latest/
MIT License
1k stars 195 forks source link

ASAP PBMC dataset, how to fine-tune? #190

Closed LiXiaoli921 closed 5 months ago

LiXiaoli921 commented 5 months ago

hi, how to use your code for the ASAP PBMC mosaic integration task? Thanks for your help.

As the file scGPT/examples /finetune_integration.py which only used the function to load data with # ## Loading and preparing data if dataset_name == "PBMC_10K": adata = scvi.data.pbmc_dataset() # 11990 × 3346 ori_batch_col = "batch" adata.obs["celltype"] = adata.obs["str_labels"].astype("category") adata.var = adata.var.set_index("gene_symbols") data_is_raw = True could you please show the cod of how you fine-tune on the mosaic ASAP PBMC dataset?

subercui commented 5 months ago

Hi, thank you for the question. You can find the multi-omics tutorial here https://github.com/bowang-lab/scGPT/blob/main/tutorials/Tutorial_Multiomics.ipynb . Also for the specific data, there is a script in our release on zenodo, https://zenodo.org/records/10466117. Please look at tutorials/Fig4_D_Multiomic_ASAP.py

subercui commented 5 months ago

Going to close the issue. Feel free to reopen if you have further questions.

LiXiaoli921 commented 5 months ago

thanks so much