frankligy
commented
8 months ago Hi @LIH-WHU,
Thanks for reaching out and nice to meet you!
I was trying to reproduce this error, so I just literally took the count matrix you shared here as the input test.full.txt, I attached below:
And then I ran the code the same as you on my end:
#!/gpfs/data/yarmarkovichlab/Frank/test_snaf/test_snaf_env/bin/python3.7
import os,sys
import pandas as pd
import numpy as np
import snaf
import anndata as ad
# change the below two paths
db_dir = '/gpfs/data/yarmarkovichlab/Frank/SNAF_ecosystem/data'
netMHCpan_path = '/gpfs/data/yarmarkovichlab/Frank/SNAF_ecosystem/netMHCpan-4.1/netMHCpan'
# initiate
df = pd.read_csv('test.full.txt',sep='\t',index_col=0)
snaf.initialize(df=df,db_dir=db_dir,binding_method='netMHCpan',software_path=netMHCpan_path)
# test something
jcmq = snaf.JunctionCountMatrixQuery(junction_count_matrix=df,cores=30,outdir='result_test')
hlas = [['HLA-A*02:01','HLA-A*02:01','HLA-B*39:10','HLA-B*15:01','HLA-C*03:03','HLA-C*12:03'],
['HLA-A*02:01','HLA-A*01:01','HLA-B*40:01','HLA-B*52:01','HLA-C*03:04','HLA-C*12:02']]
jcmq.run(hlas=hlas,outdir='result_test')
snaf.JunctionCountMatrixQuery.generate_results(path='./result_test/after_prediction.p',outdir='./result_test')But I didn't encounter any errors and all the results are generated as expected:
Current loaded gtex cohort with shape (12, 2629)
2024-03-06 09:46:14 finishing initialization
reduce valid NeoJunction from 27 to 7 because they are present in GTEx
junction_count_matrix: (27, 2)
cores: 30
valid: 7
invalid: 20
cond_df: (27, 2)
subset: (7, 2)
translated: list of 7 nj objects
cond_subset_df: (7, 2)
results: list of length None
0it [00:00, ?it/s] | 0/1 [00:00<?, ?it/s]
0it [00:00, ?it/s]
0it [00:00, ?it/s]
0it [00:00, ?it/s]
100%|███████████████████████████████████████████| 1/1 [00:00<00:00, 1285.41it/s]
100%|███████████████████████████████████████████| 1/1 [00:00<00:00, 1292.15it/s]
100%|███████████████████████████████████████████| 1/1 [00:00<00:00, 1417.95it/s]
0it [00:00, ?it/s]
0it [00:00, ?it/s]
0it [00:00, ?it/s]
0it [00:00, ?it/s]
0it [00:00, ?it/s]
0it [00:00, ?it/s]
0it [00:00, ?it/s]
0it [00:00, ?it/s]
0it [00:00, ?it/s]
0it [00:00, ?it/s]
0it [00:00, ?it/s]
0it [00:00, ?it/s]
0it [00:00, ?it/s]
0it [00:00, ?it/s]
0it [00:00, ?it/s]
0it [00:00, ?it/s]
0it [00:00, ?it/s]
0it [00:00, ?it/s]
100%|█████████████████████████████████████████████| 1/1 [00:02<00:00, 2.46s/it]
100%|█████████████████████████████████████████████| 1/1 [00:05<00:00, 5.47s/it]
Error in atexit._run_exitfuncs:
Traceback (most recent call last):
File "/gpfs/data/yarmarkovichlab/Frank/test_snaf/test_snaf_env/lib/python3.7/site-packages/snaf/dash_app/app.py", line 20, in clear_assets
imgs = os.listdir('assets')
FileNotFoundError: [Errno 2] No such file or directory: 'assets'That error itself also puzzles me a lot, because it basically says the dict_fa dictionary, which is keyed by ENSG ID and values are the sequence and other information, doesn't have the key "ENSG00000000419", but I just did a simple grep, and as you can see, it is here:
[lig08@bigpurple-ln3 yarmarkovichlab]$ cd Frank/SNAF_ecosystem/data/Alt91_db/
[lig08@bigpurple-ln3 Alt91_db]$ grep "ENSG00000000419" Hs_gene-seq-2000_flank.fa
>ENSG00000000419|20|50934867|50958555At this moment, what I can suggest is double check if the initialization step is complete or not, whether all the required database are properly downloaded and loaded into RAM. Or if it has anything to do with the environment like if you are using jupyter sometimes the global variable will be messed up if not careful. You can do the below check:
db_dir = '/gpfs/data/yarmarkovichlab/Frank/SNAF_ecosystem/data'
netMHCpan_path = '/gpfs/data/yarmarkovichlab/Frank/SNAF_ecosystem/netMHCpan-4.1/netMHCpan'
# initiate
df = pd.read_csv('test.full.txt',sep='\t',index_col=0)
snaf.initialize(df=df,db_dir=db_dir,binding_method='netMHCpan',software_path=netMHCpan_path)
print(snaf.snaf.dict_fa['ENSG00000000419']). # should be able to print out something
# or below
seq = snaf.snaf.subexon_tran(subexon='I2.1_50949267',EnsID='ENSG00000000419',flag='site1',code=False)Let me know if you can just use my script snippet to get it to run, and keep me posted, now I am very curious about what have might happened here because I can not explain it.
Best, Frank
LIH-WHU
Thank you for developing this excellent tool! I am also from Wuhan, China. Congratulatations on your remarkable achievements! I encountered the following issue while following the tutorial for 'Running the T antigen workflow' and sincerely invite your assistance. Running the following code poses no issues:
Next, proceed with the following step:
jcmq.run(hlas=hlas,outdir='./09.result')The error is as follows:
I have checked the counts.original.full.txt file. It look like this:
If I remove the line
ENSG00000000419:I2.1_50949267-I2.1_50948941 0 22the code can continue to run. However, it will encounter a similar error and get stuck at the lineENSG00000000419:I5.1_50945529-I5.1_50944551 29 59. Approximately, every 200 lines, there is one line that causes thejcmq.runto fail. The following lines also lead to similar issues:I would like to ask for your advice: Why is this error occurring? Apart from ignoring these lines (there are too many of them to delete one by one), is there a better solution? Thank you very much!