Datasets:
Tasks:
Tabular Classification
Modalities:
Text
Formats:
csv
Languages:
English
Size:
10K - 100K
License:
haneulpark
commited on
Update README.md
Browse files
README.md
CHANGED
@@ -87,6 +87,89 @@ dataset_info:
|
|
87 |
num_examples: 596
|
88 |
---
|
89 |
|
90 |
-
Human & Rat Liver Microsomal Stability
|
91 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
92 |
|
|
|
87 |
num_examples: 596
|
88 |
---
|
89 |
|
90 |
+
# Human & Rat Liver Microsomal Stability
|
91 |
+
3345 RLM and 6420 HLM compounds were initially collected from the ChEMBL bioactivity database.
|
92 |
+
(HLM ID: 613373, 2367379, and 612558; RLM ID: 613694, 2367428, and 612558)
|
93 |
+
Finally, the RLM stability data set contains 3108 compounds, and the HLM stability data set contains 5902 compounds.
|
94 |
+
For the RLM stability data set, 1542 (49.6%) compounds were classified as stable, and 1566 (50.4%) compounds were classified as unstable,
|
95 |
+
among which the training and test sets contain 2512 and 596 compounds, respectively.
|
96 |
+
The experimental data from the National Center for Advancing Translational Sciences (PubChem AID 1508591) were used as the external set.
|
97 |
+
For the HLM data set, 3799 (64%) compounds were classified as stable, and 2103 (36%) compounds were classified as unstable.
|
98 |
+
In addition, an external set from Liu et al.12 was used to evaluate the predictive power of the HLM model.
|
99 |
+
|
100 |
+
|
101 |
+
## Quickstart Usage
|
102 |
+
|
103 |
+
### Load a dataset in python
|
104 |
+
Each subset can be loaded into python using the Huggingface [datasets](https://huggingface.co/docs/datasets/index) library.
|
105 |
+
First, from the command line install the `datasets` library
|
106 |
+
|
107 |
+
$ pip install datasets
|
108 |
+
|
109 |
+
then, from within python load the datasets library
|
110 |
+
|
111 |
+
>>> import datasets
|
112 |
+
|
113 |
+
and load one of the `HLM_RLM` datasets, e.g.,
|
114 |
+
|
115 |
+
>>> HLM = datasets.load_dataset("maomlab/HLM_RLM", name = "HLM")
|
116 |
+
Downloading readme: 100%|ββββββββββββββββββββββββ| 4.40k/4.40k [00:00<00:00, 1.35MB/s]
|
117 |
+
Downloading data: 100%|ββββββββββββββββββββββββββ| 680k/680k [00:00<00:00, 946kB/s]
|
118 |
+
Downloading data: 100%|ββββββββββββββββββββββββββ| 2.11M/2.11M [00:01<00:00, 1.28MB/s]
|
119 |
+
Generating test split: 100%|βββββββββββββββββββββ| 1951/1951 [00:00<00:00, 20854.95 examples/s]
|
120 |
+
Generating train split: 100%|ββββββββββββββββββββ| 5856/5856 [00:00<00:00, 144260.80 examples/s]
|
121 |
+
|
122 |
+
and inspecting the loaded dataset
|
123 |
+
|
124 |
+
>>> HLM
|
125 |
+
HLM
|
126 |
+
DatasetDict({
|
127 |
+
test: Dataset({
|
128 |
+
features: ['NO.', 'compound_name', 'IUPAC_name', 'SMILES', 'CID', 'logBB', 'BBB+/BBB-', 'Inchi', 'threshold', 'reference', 'group', 'comments', 'ClusterNo', 'MolCount'],
|
129 |
+
num_rows: 1951
|
130 |
+
})
|
131 |
+
train: Dataset({
|
132 |
+
features: ['NO.', 'compound_name', 'IUPAC_name', 'SMILES', 'CID', 'logBB', 'BBB+/BBB-', 'Inchi', 'threshold', 'reference', 'group', 'comments', 'ClusterNo', 'MolCount'],
|
133 |
+
num_rows: 5856
|
134 |
+
})
|
135 |
+
})
|
136 |
+
|
137 |
+
### Use a dataset to train a model
|
138 |
+
One way to use the dataset is through the [MolFlux](https://exscientia.github.io/molflux/) package developed by Exscientia.
|
139 |
+
First, from the command line, install `MolFlux` library with `catboost` and `rdkit` support
|
140 |
+
|
141 |
+
pip install 'molflux[catboost,rdkit]'
|
142 |
+
|
143 |
+
then load, featurize, split, fit, and evaluate the a catboost model
|
144 |
+
|
145 |
+
import json
|
146 |
+
from datasets import load_dataset
|
147 |
+
from molflux.datasets import featurise_dataset
|
148 |
+
from molflux.features import load_from_dicts as load_representations_from_dicts
|
149 |
+
from molflux.splits import load_from_dict as load_split_from_dict
|
150 |
+
from molflux.modelzoo import load_from_dict as load_model_from_dict
|
151 |
+
from molflux.metrics import load_suite
|
152 |
+
|
153 |
+
split_dataset = load_dataset('maomlab/HLM_RLM', name = 'HLM')
|
154 |
+
|
155 |
+
split_featurised_dataset = featurise_dataset(
|
156 |
+
split_dataset,
|
157 |
+
column = "SMILES",
|
158 |
+
representations = load_representations_from_dicts([{"name": "morgan"}, {"name": "maccs_rdkit"}]))
|
159 |
+
|
160 |
+
model = load_model_from_dict({
|
161 |
+
"name": "cat_boost_classifier",
|
162 |
+
"config": {
|
163 |
+
"x_features": ['SMILES::morgan', 'SMILES::maccs_rdkit'],
|
164 |
+
"y_features": ['BBB+/BBB-']}})
|
165 |
+
|
166 |
+
model.train(split_featurised_dataset["train"])
|
167 |
+
preds = model.predict(split_featurised_dataset["test"])
|
168 |
+
|
169 |
+
classification_suite = load_suite("classification")
|
170 |
+
|
171 |
+
scores = classification_suite.compute(
|
172 |
+
references=split_featurised_dataset["test"]['BBB+/BBB-'],
|
173 |
+
predictions=preds["cat_boost_classifier::BBB+/BBB-"])
|
174 |
+
|
175 |
|