Note for: HCVpred: A web server for predicting the bioactivity of hepatitis C virus NS5B inhibitors
Note for: HCVpred: A web server for predicting the
bioactivity of hepatitis C virus NS5B inhibitors
Doi: 10.1002/jcc.26223
Overview
HCV
-
NS5B -- > one of three potential therapeutic
target HCV (other two – NS3/4A, NS5A) -- > involve in viral replication
-
Developing classification structure-activity
relationship (CSAR) model -- > for identifying substructures giving rise to
anti-HCV from 578 non-redundant compounds
-
NS5B inhibitors -- >
o
Set of 12 fingerprint descriptors -- > use as
predictive models
o
Random forest algorithm with 100 independent
data splits
-
Modelability (MODI index) of data se
o
Robust
o
0.88 exceeding – threshold 0.6
-
Predictive performance
o
Accuracy
o
Sensitivity
o
Specificity
o
Matthews coefficient
-
In-depth analysis of top 20 important
descriptors
o
Aromatic ring and alkyl side chains -- >
important for NS5B inhibition
-
Predictive model is publicly deposited -- >
allow user to predict biological activity
-
Web-server -- > use for designing more potent
and specific drugs against HCV NS5B
Extract the flow
-
Structural + chemical functions of
compounds/inhibitors and biological activity
o
Understand the relationship between physicochemical
properties on biological activity
o
Physicochemical properties of compound -- >
computed through molecular descriptors calculation software -- > obtain both
quantitative value + qualitative label
-
Data compilation and curation
o
HCV NS5B -- > ChEMBL ID: CHEMBL5375
o
1903 bioactivity data points from 1350 compounds
§
IC50 is used as bioactivity unit for further
investigate
§
865 compounds
§
Define threshold
·
<1 uM active
·
> 10 uM inactive
·
Intermediate is excluded (1-10 uM) -- > 284 inhibitors
·
Final non-redundant and curated compounds --
> 578 inhibitors
-
Fingerprint descriptors extraction
o
PaDEL-descriptor -- > get 12 molecular
fingerprints from 9 classes
§
AtomPairs 2D,
§
CDK fingerprinter,
§
CDK extended,
§
CDK graph only,
§
E-state,
§
Klekota–Roth,
§
MACCS,
§
PubChem,
§
Substructure
-
415 – active compounds, 163 – inactive compounds
-
Data set modelability
o
Similar chemical structure
§
Give rise to different bioactivity -- > not
good for the modeling
o
Similar chemical structure
§
Give rise to similar bioactivity -- > good
for the modeling
o
MODI -- > modelability index -- > telling
you whether the model is good or bad -- > this study set MODI -- >
>0.65 -- > considered to be modelable
-
Webserver
o
Using shinny app
§
User interface (ui.R)
·
Input data (SMILE string)
·
PaDEL-descritor -- > convert SMILE to
chemical fingerprint
·
Analyze through predictive model
§
Server interface (server.R)
-
Input -- > SMILES string
o
Passing through the descriptor extraction
(PaDEL)
o
Go through RF machine learning
o
Classification labeling the input SMILES, active
or inactive -- > output format .csv
Comments
Post a Comment