search

hart-lab / bagel

BAGEL software
MIT License
26 stars 20 forks source link

What if there are multiple replicates for control condition? #3

Closed rraadd88 closed 5 years ago

rraadd88 commented 5 years ago

Hi, I was trying to analyse a CRISPR screen data with multiple replicates of control condition. I have 4 replicates of control conditions (and 4 for treated/test condition). So I was using this command:

python BAGEL.py fc -i {path to input} -o {path to output} -c 1,2,3,4

Then I got this error 

Traceback (most recent call last):
  File "BAGEL.py", line 165, in <module>
    ctrl_column = int(arg)
ValueError: invalid literal for int() with base 10: '1,2,3,4'

It seems to me that bagel (in the current version) is not applicable to data with multiple replicates of control condition. Please correct me if I am wrong. Thank you. (bagel works very well if there is only one control condition.)

rraadd88 commented 5 years ago

Hi again, What if I analyse the 4 replicates of control conditions and 4 replicates for treated/test condition in a paired manner. i.e. 

1st control vs 1st test condition
2st control vs 2nd test condition
3rd control vs 3rd test condition
4th control vs 4th test condition

and then aggregate the resulting (4) foldchange vectors.

rooeikim commented 5 years ago

Hi rraadd88,

You can use 4 replicates of control when you use updated version of BAGEL. It is on the branch of this github. build_111

Thank you.

rraadd88 commented 5 years ago

Hi @rooeikim , Thank you very much for the reply. build_111 branch indeed supports multiple controls. I find bagel very useful in CRISPR screens. Great work! Thanks again.