hall-lab / svtyper

Bayesian genotyper for structural variants
MIT License
125 stars 55 forks source link

Fixed #103 (KeyError: 'CIPOS') #111

Open mchowdh200 opened 4 years ago

mchowdh200 commented 4 years ago
srbehera commented 3 years ago

I am still getting CIPOS error. Could you please check the following error?

File ".../miniconda3/envs/py2/bin/svtyper", line 11, in <module>
    load_entry_point('svtyper==0.7.1', 'console_scripts', 'svtyper')()
  File ".../miniconda3/envs/py2/lib/python2.7/site-packages/svtyper/classic.py", line 575, in cli
    sys.exit(main())
  File ".../miniconda3/envs/py2/lib/python2.7/site-packages/svtyper/classic.py", line 568, in main
    args.max_ci_dist)
  File ".../miniconda3/envs/py2/lib/python2.7/site-packages/svtyper/classic.py", line 265, in sv_genotype
    ciA = confidence_interval(var, 'CIPOS', 'CIPOS95', max_ci_dist)
  File ".../miniconda3/envs/py2/lib/python2.7/site-packages/svtyper/parsers.py", line 12, in confidence_interval
    ci = map(int, var.info[tag].split(','))
KeyError: 'CIPOS'
mchowdh200 commented 3 years ago

Hi can you provide an example vcf for me to test with? I'll try to see what I can do.

srbehera commented 3 years ago

Sure. Here is the VCF file I am using https://bcm.box.com/s/otlv5zqdc0nbkgg1l4phskeelbfgrry9

Please let me know if you cannot download that file.

srbehera commented 3 years ago

It works good by adding "CIPOS=-100,100;CIEND=-100,100" to each SV entry in vcf file.

dr-ashu-geno commented 2 years ago

Dear @srbehera

Thank you for commenting on this post; I added "CIPOS=-100,100;CIEND=-100,100" to all my records and it worked; however, I'm worried about the genotyping results. I mean, doesn't adding "CIPOS=-100,100;CIEND=-100,100" to the variants affect the genotyping accuracy in samples?

Thank you in advance for your time, Best regards,

srbehera commented 2 years ago

Hello @dr-ashu-geno Sorry for my late response. I am not sure if it affected the accuracy of our analysis. I will take a look at the outputs and let you know if I find anything.