issues
search
maxhodak
/
keras-molecules
Autoencoder network for learning a continuous representation of molecular structures.
MIT License
519
stars
146
forks
source link
issues
Newest
Newest
Most commented
Recently updated
Oldest
Least commented
Least recently updated
Preprocess uses too much RAM
#35
dakoner
opened
7 years ago
21
Reproduce GDB-17 construction pipeline
#34
dakoner
opened
7 years ago
4
Strange results for the autoencoder with preloaded weights
#33
jeammimi
opened
7 years ago
1
add script to convert rdkit mols to networkx graphs with attributes
#32
dakoner
opened
7 years ago
7
Better molecule representation
#31
maxhodak
closed
1 year ago
17
Make interpolate act like a library.
#30
dakoner
closed
7 years ago
0
Script to interpolate between two SMILES strings using trained network
#29
dakoner
closed
7 years ago
0
use https rather than git:// in requirement.txt
#28
hsiaoyi0504
closed
7 years ago
2
Trained model
#27
hsiaoyi0504
closed
7 years ago
1
problem with loading sample
#26
jeammimi
closed
7 years ago
7
sample: added --smiles and --save_json options
#25
dribnet
opened
7 years ago
7
Added missing import to sample.py
#24
dribnet
closed
7 years ago
1
Validation dataset
#23
alainrichardt
closed
7 years ago
2
Add script to download ZINC clean drug-like SMILES and process it.
#22
dakoner
closed
7 years ago
0
Decoder network and overhauled sampling script
#21
maxhodak
closed
7 years ago
0
fix Lambda layer in autoencoder to work with Theano, correct branch
#20
pechersky
closed
7 years ago
1
Chunk adding tensor to h5 file to prevent out_of_memory errors
#19
pechersky
closed
7 years ago
3
Trying to preprocess an h5 file with a large number of structures (~500K) results in out-of-memory error
#18
pechersky
closed
7 years ago
1
change sample.py to take a input data file instead of assuming a default
#17
pechersky
closed
7 years ago
1
sample.py assumes what input file to sample from
#16
pechersky
closed
7 years ago
1
Change train and test separation to respect indexing of structures
#15
pechersky
closed
7 years ago
1
h5 files with non-sequential indices cannot be processed
#14
pechersky
closed
7 years ago
0
LogP calculations for 50k dataset
#13
tantrev
opened
7 years ago
4
fix Lambda layer in autoencoder to work with Theano
#12
pechersky
closed
7 years ago
1
Property prediction
#11
dribnet
opened
7 years ago
15
Sharing trained weight
#10
jeammimi
closed
7 years ago
4
Script to download ChEMBL SMILES entries and convert them to h5 input…
#9
dakoner
closed
7 years ago
2
Make preprocessing command line reference included example.
#8
dakoner
closed
7 years ago
0
Add an initial Dockerfile.
#7
dakoner
closed
7 years ago
0
Add an initial Dockerfile.
#6
dakoner
closed
7 years ago
0
Add an initial Dockerfile.
#5
dakoner
closed
7 years ago
0
Architecture issues
#4
fchollet
closed
7 years ago
7
Error: No object named table in the file
#3
paularindam
closed
7 years ago
2
Open source license?
#2
shoyer
closed
7 years ago
2
Use timedistributed wrapper
#1
farizrahman4u
closed
7 years ago
2
Previous