fimrie / deepcoy Goto Github PK
View Code? Open in Web Editor NEWLicense: Other
License: Other
Hello There,
I work on MacOSX Catalina with Anaconda.
after I ran
conda env create --file DeepCoy-env.yml
I received the following error:
Collecting package metadata (repodata.json): done
Solving environment: failed
ResolvePackageNotFound:
- gstreamer==1.14.0=hb453b48_1
- icu==58.2=h9c2bf20_1
- grpcio==1.16.1=py36hf8bcb03_1
- openssl==1.1.1a=h7b6447c_0
- bzip2==1.0.6=h14c3975_5
- pillow==5.3.0=py36h34e0f95_0
- pyqt==5.9.2=py36h22d08a2_1
- rdkit==2018.09.1.0=py36h71b666b_1
- cython==0.29=py36he6710b0_0
- libgfortran-ng==7.3.0=hdf63c60_0
- cudnn==7.2.1=cuda9.2_0
- mkl_fft==1.0.6=py36h7dd41cf_0
- sqlite==3.25.3=h7b6447c_0
- pandas==0.23.4=py36h04863e7_0
- protobuf==3.6.1=py36he6710b0_0
- jpeg==9b=h024ee3a_2
- tensorflow-gpu==1.10.0=hf154084_0
- libpng==1.6.35=hbc83047_0
- pixman==0.34.0=hceecf20_3
- py-boost==1.65.1=py36hf484d3e_4
- tensorflow-base==1.10.0=gpu_py36had579c0_0
- numpy==1.15.4=py36h1d66e8a_0
- libstdcxx-ng==8.2.0=hdf63c60_1
- zlib==1.2.11=h7b6447c_3
- mkl_random==1.0.1=py36h4414c95_1
- xz==5.2.4=h14c3975_4
- cupti==9.2.148=0
- dbus==1.13.12=h746ee38_0
- readline==7.0=ha6073c6_4
- libgcc-ng==8.2.0=hdf63c60_1
- libedit==3.1.20170329=h6b74fdf_2
- c-ares==1.15.0=h7b6447c_1
- cudatoolkit==9.2=0
- gst-plugins-base==1.14.0=hbbd80ab_1
- numpy-base==1.15.4=py36h81de0dd_0
- tk==8.6.8=hbc83047_0
- cairo==1.14.12=h8948797_3
- libprotobuf==3.6.1=hd408876_0
- libxml2==2.9.8=h26e45fe_1
- ncurses==6.1=hf484d3e_0
- python==3.6.7=h0371630_0
- libxcb==1.13=h1bed415_1
- scipy==1.1.0=py36hfa4b5c9_1
- fontconfig==2.13.0=h9420a91_0
- libffi==3.2.1=hd88cf55_4
- libboost==1.65.1=habcd387_4
- expat==2.2.6=he6710b0_0
- freetype==2.9.1=h8a8886c_1
- libuuid==1.0.3=h1bed415_2
- tensorflow==1.10.0=gpu_py36hcebf108_0
- scikit-learn==0.20.1=py36h4989274_0
- _tflow_1100_select==0.0.1=gpu
- tensorboard==1.10.0=py36hf484d3e_0
- kiwisolver==1.1.0=py36he6710b0_0
- tornado==6.0.3=py36h7b6447c_0
- glib==2.56.2=hd408876_0
- matplotlib==3.0.2=py36h5429711_0
- libtiff==4.0.9=he85c1e1_2
- sip==4.19.13=py36he6710b0_0
- qt==5.9.7=h5867ecd_1
- pcre==8.42=h439df22_0
What could be the possible solution to this problem?
Thanks in advance
I encountered the following error: “FileNotFoundError: [Errno 2] No such file or directory: ‘…/models/DeepCoy_DUDE_model_e09.pickle’”, and I cannot find the path “models”. Could you please suggest a solution?
Hi I'm found that your model atom dictionary doesn't have phosphorus, and I added it. but I didn't modify bucket_sizes, because I don't know what it does
elif dataset=='zinc':
return { 'atom_types': ['Br1(0)', 'C4(0)', 'Cl1(0)', 'F1(0)', 'H1(0)', 'I1(0)',
'N2(-1)', 'N3(0)', 'N4(1)', 'O1(-1)', 'O2(0)', 'S2(0)','S4(0)', 'S6(0)', 'P5(0)'],
'maximum_valence': {0: 1, 1: 4, 2: 1, 3: 1, 4: 1, 5:1, 6:2, 7:3, 8:4, 9:1, 10:2, 11:2, 12:4, 13:6, 14:5},
'number_to_atom': {0: 'Br', 1: 'C', 2: 'Cl', 3: 'F', 4: 'H', 5:'I', 6:'N', 7:'N', 8:'N', 9:'O', 10:'O', 11:'S', 12:'S', 13:'S', 14:'P'},
'bucket_sizes': np.array([28,31,33,35,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,53,55,58,84])
}
BUT failed when training. Is there anything else that needs to be modified?
Can you help me? Thanks!
Hello!
I want to use DeepCoy for decoy generation based on 3172 active structures obtained by chembl. I successfully installed DeepCoy and generated output decoys for the P38-alpha_example from the jupyter notebook and for a small subset of my own actives (10 first ligands from my 3062 actives).
Nonetheless when I try to generate decoys for all 3062 of my own actives I get an Index error.
IndexError: index 28 is out of bounds for axis 1 with size 28
All my input files, commands and terminal output including the error are available at: https://box.fu-berlin.de/s/6WiPi8ejTkJpysD
I know that numpy starts counting at 0 and that 28 is out of the range of 0-27 but I do not know why this error occurs only in a larger subset as I do not see any differences in the file formats of the input files of the 10 or the 3062 ligands. I also ensured that both files (.smi / .json) contain the same number of ligands (i.e. I ensured, that all molecules from the .smi could be red and processed by prepare_data.py to create the .json file. ).
I hope you can help me to solve the issue
Hi,
When running the examples to generate molecules with the pretrained models, or running the example Jupyter Notebook (both in a newly made conda environment setup with the DeepCoy-env.yml file) I get a tensorflow error:
Input:
(DeepCoy-env) XXXX@XXXX:~/DeepCoy$ python3 DeepCoy.py --restore models/DeepCoy_DUDE_model_e09.pickle --dataset zinc --config '{"generation": true, "number_of_generation_per_valid": 1000, "batch_size": 1, "train_file": "data/molecules_zinc_decoys_valid.json", "valid_file": "data/molecules_zinc_dekois_valid.json", "output_name": "output/DeepCoy_generated_decoys_zinc_dekois_valid.txt"}'
Output:
Traceback (most recent call last):
File "DeepCoy.py", line 1129, in <module>
model = DenseGGNNChemModel(args)
File "DeepCoy.py", line 55, in __init__
super().__init__(args)
File "/DeepCoy/GGNN_DeepCoy.py", line 80, in __init__
self.make_model()
File "/DeepCoy/GGNN_DeepCoy.py", line 126, in make_model
self.prepare_specific_graph_model()
File "DeepCoy.py", line 196, in prepare_specific_graph_model
cell = tf.nn.rnn_cell.DropoutWrapper(cell,
AttributeError: module 'tensorflow.python.ops.nn' has no attribute 'rnn_cell'
When trying to debug, it looks as though tf.nn.rnn_cell was moved into tf.contrib from version 1.0 (this contrib usage seems to be used in line 195 and 209 of DeepCoy also?)
Changing lines 195, 196 and 209, 210 from:
cell = tf.contrib.rnn.GRUCell(new_h_dim)
cell = tf.nn.rnn_cell.DropoutWrapper(cell,
to
cell = tf.contrib.rnn.GRUCell(new_h_dim)
cell = tf.contrib.rnn.DropoutWrapper(cell,
Fixes this particular error, but then the below error is produced from the same input:
Output:
File "DeepCoy.py", line 1129, in <module>
model = DenseGGNNChemModel(args)
File "DeepCoy.py", line 55, in __init__
super().__init__(args)
File "/DeepCoy/GGNN_DeepCoy.py", line 80, in __init__
self.make_model()
File "/DeepCoy/GGNN_DeepCoy.py", line 142, in make_model
"_encoder")
File "DeepCoy.py", line 307, in compute_final_node_representations_with_residual
h = self.weights['node_gru'+scope_name+str(iter_idx)](acts, h)[1] # [b*v, h]
File "/.local/lib/python3.6/site-packages/tensorflow/contrib/rnn/python/ops/core_rnn_cell_impl.py", line 713, in __call__
output, new_state = self._cell(inputs, state, scope)
File /.local/lib/python3.6/site-packages/tensorflow/contrib/rnn/python/ops/core_rnn_cell_impl.py", line 150, in __call__
[inputs, state], 2 * self._num_units, True, 1.0))
File "/.local/lib/python3.6/site-packages/tensorflow/contrib/rnn/python/ops/core_rnn_cell_impl.py", line 1044, in _linear
_WEIGHTS_VARIABLE_NAME, [total_arg_size, output_size], dtype=dtype)
File "/.local/lib/python3.6/site-packages/tensorflow/python/ops/variable_scope.py", line 1049, in get_variable
use_resource=use_resource, custom_getter=custom_getter)
File "/.local/lib/python3.6/site-packages/tensorflow/python/ops/variable_scope.py", line 948, in get_variable
use_resource=use_resource, custom_getter=custom_getter)
File "/.local/lib/python3.6/site-packages/tensorflow/python/ops/variable_scope.py", line 356, in get_variable
validate_shape=validate_shape, use_resource=use_resource)
File "/.local/lib/python3.6/site-packages/tensorflow/python/ops/variable_scope.py", line 341, in _true_getter
use_resource=use_resource)
File "/.local/lib/python3.6/site-packages/tensorflow/python/ops/variable_scope.py", line 653, in _get_single_variable
name, "".join(traceback.format_list(tb))))
ValueError: Variable graph_model/gru_scope_encoder0/gru_cell/gates/weights already exists, disallowed. Did you mean to set reuse=True in VarScope? Originally defined at:
File "/.local/lib/python3.6/site-packages/tensorflow/contrib/rnn/python/ops/core_rnn_cell_impl.py", line 1044, in _linear
_WEIGHTS_VARIABLE_NAME, [total_arg_size, output_size], dtype=dtype)
File "/.local/lib/python3.6/site-packages/tensorflow/contrib/rnn/python/ops/core_rnn_cell_impl.py", line 150, in __call__
[inputs, state], 2 * self._num_units, True, 1.0))
File "/.local/lib/python3.6/site-packages/tensorflow/contrib/rnn/python/ops/core_rnn_cell_impl.py", line 713, in __call__
output, new_state = self._cell(inputs, state, scope)
> /.local/lib/python3.6/site-packages/tensorflow/python/ops/variable_scope.py(653)_get_single_variable()
-> name, "".join(traceback.format_list(tb))))
Hi,
In your article, you mentioned that you compute 27 unique properties.
But, in your script decoy_utils.py calc_props_all
function, I only find 26 properties. I checked many times, seems like it is 26 properties.
Any suggestions? thanks!
hello! when I use the prepare_data.py to process my own training dataset, I got an error below:
Range Error
idx
Violation occurred on line 168 in file /opt/conda/conda-bld/rdkit_1540176401003/work/Code/GraphMol/ROMol.cpp
Failed Expression: 31 < 31
I found this error when I tried to generate decoys
AttributeError: module 'tensorflow' has no attribute 'ConfigProto'
Hi,
Thanks for the amazing work done here, I was wondering if there was a plan to upgrade to python 3.10 (which If I understand correctly would require moving to TensorFlow 2.x)?
Cheers,
Tony
A declarative, efficient, and flexible JavaScript library for building user interfaces.
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google ❤️ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.