amina01 / acranker Goto Github PK

Hi! Thank you for your repo!
I am trying to train XGBRanker in xgbranker.py.
With xgboost==0.90, it reports

File "AcRanker/xgbranker.py", line 108, in fit
    xgb_model=None)
xgboost.core.XGBoostError: Invalid Parameter format for seed expect int but value='None'

With xgboost==1.0.0, it reports

File "AcRanker/xgbranker.py", line 108, in fit
    xgb_model=None)
xgboost.core.XGBoostError: Invalid Parameter format for verbosity expect int but value='True'

I have searched on google but got no solution. Could you please me? I will appreciate.

Hi there,

When I try to use the AcRanker tool through Conda, I receive the following error:

Traceback (most recent call last): File "/home/ubuntu/AcRanker/acranker.py", line 9, in <module> from server2 import * File "/home/ubuntu/AcRanker/server2.py", line 29, in <module> from sklearn.externals import joblib ImportError: cannot import name 'joblib' from 'sklearn.externals' (/opt/miniconda3/envs/AcRanker/lib/python3.11/site-packages/sklearn/externals/__init__.py) (AcRanker) ubuntu@jdtest2:~/AcRanker$ python acranker.py --help Traceback (most recent call last): File "/home/ubuntu/AcRanker/acranker.py", line 9, in <module> from server2 import * File "/home/ubuntu/AcRanker/server2.py", line 29, in <module> from sklearn.externals import joblib ImportError: cannot import name 'joblib' from 'sklearn.externals' (/opt/miniconda3/envs/AcRanker/lib/python3.11/site-packages/sklearn/externals/__init__.py)

I was wondering if anyone has come across this error and if so, how they resolved it?

Thanks in advance!

Hi,
I used your online tool and got a couple of proteins with scores. What is an acceptable score for an Acr hit? What do positive and negative scores imply?

Thanks

Hi,

We are currently planning to use Acranker for a portion of our project. As new Anti-CRISPR proteins were experimentally verified and published since the original training of Acranker (a total of 57 published Anti-CRISPRs as of now), would it be possible for you to update the model to include the latest discovered Anti-CRISPR proteins? Thanks!

Best,
Bowen

I just ran AcRanker on 339 known Acrs from AcrHub. Only 22 of the 339 proteins had positive scores (6.5%). Besides using the rank, is there a recommended score threshold for classifying Acrs?

Edit: just saw you already answered this

Hi,

When trying to run the Acranker locally, the following error came up:

Traceback (most recent call last):
File "acranker.py", line 14, in
analyzePhage(prot_path, rank_path)
File "/mnt/array1/bowen/TOOLS/AcRanker/server2.py", line 237, in analyzePhage
model=pickle.load(open('xgb_rank.pickle', 'rb'))
File "/home/yangbowen/.local/lib/python3.6/site-packages/xgboost/core.py", line 1093, in setstate
_LIB.XGBoosterUnserializeFromBuffer(handle, ptr, length))
File "/home/yangbowen/.local/lib/python3.6/site-packages/xgboost/core.py", line 189, in _check_call
raise XGBoostError(py_str(_LIB.XGBGetLastError()))

I believe this is a problem with the version of xgboost being used. I have the '1.0.0' version. could you help me bypass this error?