Installation Failure
NiklausChung opened this issue · 5 comments
Hi, there!
I've downloaded 'DeepBGC' on my Mac successfully, but there's some issue reported as follows when I tapped in 'deepbgc info'.
=================|_|===== version 0.1.30 ===== INFO 26/06 22:03:07 Available data files: ['Pfam-A.31.0.hmm.h3f', 'Pfam-A.31.0.hmm.h3i', 'Pfam-A.31.0.clans.tsv', 'Pfam-A.31.0.hmm', 'Pfam-A.31.0.hmm.h3p', 'Pfam-A.31.0.hmm.h3m'] INFO 26/06 22:03:07 ================================================================================ INFO 26/06 22:03:07 Available detectors: ['clusterfinder_retrained', 'clusterfinder_geneborder', 'clusterfinder_original', 'deepbgc'] INFO 26/06 22:03:07 -------------------------------------------------------------------------------- INFO 26/06 22:03:07 Model: clusterfinder_retrained INFO 26/06 22:03:07 Loading model from: /Users/Work/Library/Application Support/deepbgc/data/0.1.0/detector/clusterfinder_retrained.pkl WARNING 26/06 22:03:07 Model not supported: ('Package "hmmlearn" needs to be installed to run ClusterFinder HMM. ', 'Install extra dependencies using: \n pip install "deepbgc[hmm]"') INFO 26/06 22:03:07 -------------------------------------------------------------------------------- INFO 26/06 22:03:07 Model: clusterfinder_geneborder INFO 26/06 22:03:07 Loading model from: /Users/Work/Library/Application Support/deepbgc/data/0.1.0/detector/clusterfinder_geneborder.pkl WARNING 26/06 22:03:07 Model not supported: ('Package "hmmlearn" needs to be installed to run ClusterFinder HMM. ', 'Install extra dependencies using: \n pip install "deepbgc[hmm]"') INFO 26/06 22:03:07 -------------------------------------------------------------------------------- INFO 26/06 22:03:07 Model: clusterfinder_original INFO 26/06 22:03:07 Loading model from: /Users/Work/Library/Application Support/deepbgc/data/0.1.0/detector/clusterfinder_original.pkl WARNING 26/06 22:03:07 Model not supported: ('Package "hmmlearn" needs to be installed to run ClusterFinder HMM. ', 'Install extra dependencies using: \n pip install "deepbgc[hmm]"') INFO 26/06 22:03:07 -------------------------------------------------------------------------------- INFO 26/06 22:03:07 Model: deepbgc INFO 26/06 22:03:07 Loading model from: /Users/Work/Library/Application Support/deepbgc/data/0.1.0/detector/deepbgc.pkl Using TensorFlow backend. WARNING 26/06 22:03:08 Model not supported: ("Error unpickling model from path '/Users/Work/Library/Application Support/deepbgc/data/0.1.0/detector/deepbgc.pkl'", TypeError('Descriptors cannot not be created directly.\nIf this call came from a _pb2.py file, your generated code is out of date and must be regenerated with protoc >= 3.19.0.\nIf you cannot immediately regenerate your protos, some other possible workarounds are:\n 1. Downgrade the protobuf package to 3.20.x or lower.\n 2. Set PROTOCOL_BUFFERS_PYTHON_IMPLEMENTATION=python (but this will use pure-Python parsing and will be much slower).\n\nMore information: https://developers.google.com/protocol-buffers/docs/news/2022-05-06#python-updates')) INFO 26/06 22:03:08 ================================================================================ INFO 26/06 22:03:08 Available classifiers: ['product_activity', 'product_class'] INFO 26/06 22:03:08 -------------------------------------------------------------------------------- INFO 26/06 22:03:08 Model: product_activity INFO 26/06 22:03:08 Loading model from: /Users/Work/Library/Application Support/deepbgc/data/0.1.0/classifier/product_activity.pkl /Users/Work/opt/miniconda3/envs/deepbgc/lib/python3.7/site-packages/sklearn/base.py:306: UserWarning: Trying to unpickle estimator DecisionTreeClassifier from version 0.18.2 when using version 0.21.3. This might lead to breaking code or invalid results. Use at your own risk. UserWarning) /Users/Work/opt/miniconda3/envs/deepbgc/lib/python3.7/site-packages/sklearn/base.py:306: UserWarning: Trying to unpickle estimator RandomForestClassifier from version 0.18.2 when using version 0.21.3. This might lead to breaking code or invalid results. Use at your own risk. UserWarning) INFO 26/06 22:03:08 Type: RandomForestClassifier INFO 26/06 22:03:08 Version: 0.1.0 INFO 26/06 22:03:08 Timestamp: 1551781433.886473 (2019-03-05T18:23:53.886473) INFO 26/06 22:03:08 -------------------------------------------------------------------------------- INFO 26/06 22:03:08 Model: product_class INFO 26/06 22:03:08 Loading model from: /Users/Work/Library/Application Support/deepbgc/data/0.1.0/classifier/product_class.pkl INFO 26/06 22:03:08 Type: RandomForestClassifier INFO 26/06 22:03:08 Version: 0.1.0 INFO 26/06 22:03:08 Timestamp: 1551781410.019103 (2019-03-05T18:23:30.019103) INFO 26/06 22:03:08 ================================================================================ WARNING 26/06 22:03:08 Some warnings detected, check the output above
It looks like something is blocking me from accessing the file 'deepbgc.pkl', or maybe something is incompatible with 'deepbgc.pkl'. I tried to run the script as stated in 'README', using a partial sequence from my genome. It seems the question stands that 'deepbgc.pkl' is not working as supposed (see the picture).
)
Hi @NiklausChung thanks for reporting this. We fixed the protobuf dependency version, please try reinstalling deepbgc into a new environment. Or this might do the same trick: conda update deepbgc
Hi@prihoda thanks for your help. After reinstalling deepbgc as you recommended, now deepbgc is fully functional on my mac.
However, the whole 'sideload json file to AntiSMASH' thing still bothers me. In short, deepbgc predicted that there's over 300 BGCs in the genome of interest (see the quote), which is pretty impressive given the rare nature of the species; but when I uploaded the genome file and the extra annotation (aka. deepbgc's json file, see the picture), the result was reduced to only 18 BGCs (see the picture). Considering that without the extra json file provided to the AntiSMASH, the result was no different. I was wondering that if I used the json file correctly, and if you could help me understand how the result can be sideload to antiSMASH.
================================================================================
Detected 306 total BGCs using deepbgc model
Number of BGCs with predicted product_class:
no confident class: 199
Polyketide: 69
Terpene: 30
RiPP: 12
NRP: 9
Saccharide: 2
Other: 1
Number of BGCs with predicted product_activity:
antibacterial: 216
no confident class: 80
antifungal: 18
inhibitor: 10
========================================
Saved DeepBGC result to: Assembled_Genome
========================================
@prihoda I actually installed DeepBGC today as guided by creating a new conda environment and followed each step, however, when I tried deepbgc pipeline contig.fasta, I encountered an error due to protobuf package as attached in screenshot. Would you please advise?
e
@prihoda I actually installed DeepBGC today as guided by creating a new conda environment and followed each step, however, when I tried
deepbgc pipeline contig.fasta, I encountered an error due toprotobufpackage as attached in screenshot. Would you please advise?
e
I think you should try to downgrade the buffer to 3.19.0 (worked for me). Just do as the last lines in screenshot recommended.
Should now be fixed in new deepbgc version on bioconda