ddl.pp.calculate_threshold issue about dandelion HOT 6 CLOSED

Hi, thank you for using the package. My apologies for the bug. I’ve picked that up and it’s has been fixed in the devel branch. Do you want to try and install from there and see if it works? The instructions to install fro the devel branch should be on the main readme.

from dandelion.

Thank you so much for your fast response. I believe by devel branch you meant this https://github.com/zktuong/dandelion/blob/devel/README.md but I don't see any difference. The only different thing I did now was to install sc-dandelion with the branch option pip install git+https://github.com/zktuong/dandelion.git@devel. I get a new error.

ddl.pp.calculate_threshold(vdj, manual_threshold = 0.1)
RRuntimeError Traceback (most recent call last)
/var/folders/3x/r5d2xyzx39sgl5nq6cj8fmvw0000gn/T/ipykernel_22663/3386437591.py in
----> 1 ddl.pp.calculate_threshold(vdj, manual_threshold = 0.1)

~/miniconda3/envs/dandelion/lib/python3.7/site-packages/dandelion/preprocessing/_preprocessing.py in calculate_threshold(self, mode, manual_threshold, VJthenLen, onlyHeavy, model, normalize_method, threshold_method, edge, cross, subsample, threshold_model, cutoff, sensitivity, specificity, ncpu, plot, plot_group, figsize, **kwargs)
2594 warnings.filterwarnings("ignore")
2595
-> 2596 sh = importr('shazam')
2597 pandas2ri.activate()
2598 if 'v_call_genotyped' in dat.columns:

~/miniconda3/envs/dandelion/lib/python3.7/site-packages/rpy2/robjects/packages.py in importr(name, lib_loc, robject_translations, signature_translation, suppress_messages, on_conflict, symbol_r2python, symbol_resolve, data)
482 if _package_has_namespace(name,
483 _system_file(package=name)):
--> 484 env = _get_namespace(name)
485 version = _get_namespace_version(name)[0]
486 exported_names = set(_get_namespace_exports(name))

~/miniconda3/envs/dandelion/lib/python3.7/site-packages/rpy2/rinterface_lib/conversion.py in _(*args, **kwargs)
43 def _cdata_res_to_rinterface(function):
44 def _(*args, **kwargs):
---> 45 cdata = function(*args, **kwargs)
46 # TODO: test cdata is of the expected CType
47 return _cdata_to_rinterface(cdata)

~/miniconda3/envs/dandelion/lib/python3.7/site-packages/rpy2/rinterface.py in call(self, *args, **kwargs)
678 )
679 if error_occured[0]:
--> 680 raise embedded.RRuntimeError(_rinterface._geterrmessage())
681 return res
682

RRuntimeError: loadNamespace(j <- i[[1L]], c(lib.loc, .libPaths()), versionCheck = vI[[j]]) でエラー:
‘Biostrings’ という名前のパッケージはありません

from dandelion.

The new error is related to shazam not installed, or rpy2 not pointing to the right R.

Can I ask where did you install the various R dependencies listed in the installation instructions?

in python, if you run this:

from rpy2.robjects.packages import importr
base = importr('base')
print(base._libPaths())

Is shazam installed in any of the libPaths?

from dandelion.

from rpy2.robjects.packages import importr
base = importr('base')
print(base._libPaths())

[1] "/Library/Frameworks/R.framework/Versions/4.1/Resources/library"

from dandelion.

ok you need to run the following in R separately:

install.packages(c("optparse", "alakazam", "tigger", "airr", "shazam"))

and note where they are installed.

when you run the rpy2 code chunk above, those R dependencies should be in the same path.

For instructions on how to install the immcantation pacakges, you can refer to

https://alakazam.readthedocs.io/en/stable/install/
https://shazam.readthedocs.io/en/stable/install/
https://tigger.readthedocs.io/en/stable/install/

If after installing and your python is still not talking to the right R, i would suggest to save your AIRR table, follow the tutorial on this page:
https://shazam.readthedocs.io/en/stable/vignettes/DistToNearest-Vignette/

and then run changeo's DefineClones.py
https://changeo.readthedocs.io/en/stable/examples/cloning.html

Then you can read the file back into dandelion for downstream processing.

To save the data as an AIRR table, you can do this:

# assuming vdj is the dandelion object
vdj.write_airr('airr_table.tsv')

Then to read the AIRR table back in:

vdj2 = ddl.Dandelion('new_airr_table.tsv')

Hopefully that makes sense.

from dandelion.

Thanks. I was able to install it properly and it is working ok now.

from dandelion.