Closed romunov closed 7 months ago
Hi @romunov,
Yes, your understanding is correct, you would only have to exclude any TF-gene interaction exclusively derived from SIGNOR. Thus, as you already described, your example would not need to be removed.
We're currently also working on incorporating this in our functions by providing the option to filter CollecTRI for non-academic use. I will make sure to let you know as soon as it becomes available.
Best regards Sophia
It would seem this can also be done:
import decoupler as dc
from omnipath.constants import License
ct = dc.get_collectri(organism="human", license=License.COMMERCIAL)
List of licenses can be found here: https://github.com/saezlab/omnipath/blob/main/omnipath/constants/_constants.py#L69
Thanks @romunov! SIGNOR also just changed its licence to CC BY (https://signor.uniroma2.it/, see news section) so there is not longer a need to filter it out.
Best regards Sophia
This is so easy to miss, thank you for pointing it out!
To avoid using any unlicensed resources, I would like to filter out the SIGNOR entries as noted in the license of CollecTRI repository:
Looking at the resource file, conveniently uploaded to Zenodo (#7 thanks!), it's not entirely clear to me how to filter these entries out. I would really appreciate if you could confirm my suspicion below.
For example, entry
lists
CollecTRI;SIGNOR_CollecTRI;TRRUST_CollecTRI
(TRRUST as part of DoRothEA) as resources. Am I correct to assume that source target pair of NFKB1 -> NPPB appears in SIGNOR and TRRUST? Going a step further, would this mean that this entry can be retained if SIGNOR resources are being omitted? If yes, then entries that have only SIGNOR as resource should be omitted?