In silico software used for

prediction of Genotoxicity and

mutagenicity

In Silico
Softwares
Mr. Sagar K. Savale
M. Pharm
Department of Pharmaceutics

Sagar K. Savale
 Table 1: In Silico software: Method, Prediction, Applicability and Availability
Name of Method Prediction Applicability Availability
software
TOPKAT QSAR statistical Toxicolical end Optimum No Freely
method on 2D point,include predictive space available
descriptors (e- mutagenicity,devlopment
state, topological) al toxicitiy
rodent,carcinogenicity
VEGA Caesar QSAR statistical Mutagenicity(ames),carci Quantitative AD Freely
model based on nogenicity,developmenta Measurement available
Neural Network + l toxicity,skin
rule-based model sensitization,bioconcentr
ation factor
+Toxtree Collection of Carcinogenicity, Not applicable Freely
knowledge rule mutagenicity, aromatic available
Based amine, alpha, beta,
unsaturated aldehyde.
ACD Tox Suite Statistical model Ames genotoxicity Reliability index No Freely
employing available
binominal PLS,
predefined set of
fragmental
descriptors,
local correction to
baseline, using
experimental data
for similar
compound
VEGA SARpy Statistical model Ames mutagenicity Quantitative AD Freely
based on Measurement available
structural alerts
automatically
selected on the
basis of their
occurrence in
toxic/nontoxic
compounds
Derek Nexus Collection of developmental toxicity (3 Not applicable No Freely
knowledgerule alerts), teratogenicity (5 available
based alerts), testicular toxicity
(1 alert) and
oestrogenicity (4 alerts)
Leadscope QSAR The Leadscope software Classification Commercial
methodologies has a module containing models for
QSAR models for developmental
predicting the toxicity in the
developmental toxicity of rodent fetus
the rodent foetus, dysmorphogenes
including is (structural and
 dysmorphogenesis visceral birth
(structural and visceral defects),
birth defects), developmental
developmental toxicity toxicity (fetal
(foetal growth retardation growth
and weight decrease), retardation and
and foetal survival weight
(Foetal death, post- decrease), and
implantation loss, and fetal survival
preimplantation loss). (fetal death,
The Leadscope QSAR post-
models for reproductive implantation
toxicity include rodent loss,
male reproductive, rodent Preimplantation
male sperm, female loss). Models of
reproductive reproductive
toxicity: rodent
male
reproductive,
rodent male
sperm, female
reproductive
Endocrine In vitro or in vivo Computer based Quantitative Commercial
Disruptor tests. predictive models to models to
Knowledge in silico predict the binding predict the
Base (EDKB) models for affinity of compounds to binding affinity
database (US endocrine the oestrogen and of compounds to
FDA) disruption androgen the estrogen and
nuclear receptor proteins androgen
nuclear receptor
proteins
Multicase based on a quantitative prediction Classification Commercial
(MC4PC) fragment-based of mutagenicity which is models for
technology then further refined developmental
sometimes through taking into toxicity
referred to as the consideration physico- associated with
CASE approach chemical properties as a variety of
well as the existence of Datasets, mainly
potential deactivating drugs. The
fragments marketed
or biophobes software
includes
modules for
predicting
mammal sperm
toxicity,
developmental
toxicity,
developmental
fetal growth
 retardation,
development
fetal weight
decrease and
survival fetal
death
ADMET Neural network This commercial Qualitative and Commercial
Predictor ensemble program is designed to quantitative
estimate certain ADMET prediction of
(Absorption, oestrogen
Distribution, receptor toxicity
Metabolism, Elimination, in rats.
and Toxicity) properties Based on two
of a drug-like chemical models: a
from its molecular qualitative
structure. It includes a model and, if
qualitative assessment of toxic, the
oestrogen receptor quantitative
toxicity in rats ratio of IC50
(TOX_ER_filter), estradiol/IC50
together with a compound).
quantitative measure of Molecular
oestrogen receptor descriptor space
toxicity in rats (TOX_ER
(IC50(estrogen)) that is
applied only for
compounds classified as
Toxic by the previous
model
MolCode QSAR predicting toxicological Quantitative Commercial
Toolbox methodologies endpoints and ADME prediction of rat
properties between them ER binding
endocrine activity affinity and AhR
binding affinity
OSIRIS QSAR Classification Freely
property methodologies model which available
explorer predicts
undesirable
effects
(mutagenicity,
tumorigenicity,
irritating effects
and
reproductive
effects), mainly
based on
the RTECS
database of
>3500
compounds
 PASS QSAR Prediction of Activity Classification Commercial
methodologies Spectra for Substances models giving
The system predicts the probability of
probability (Pa) of a reprotoxic
biological activity effects. The
for a new compound, by embryotoxicity
estimating the model predicts
similarity/dissimilarity of the probability
the new substance to that a substance
substances with well- crosses the
known biological placental
activities present in the membrane and
training set causes any toxic
effect (e.g. fetal
bradycardia, low
Birth weight) or
death of an
embryo. The
teratogenicity
model predicts
the probability
that a substance
crosses the
placental
membrane and
causes
abnormal
development of
one or more
body systems in
the embryo
T.E.S.T.: The QSAR It estimates the toxicity Quantitative AD Freely
Toxicity methodologies of a compound by Measurement available
Estimation Consensus applying several QSAR Developmental
Software Tool method: average methodologies thus toxicity
of the predicted allowing the user to have estimation. The
toxicity values by greater confidence in prediction is
3 models: neural predicted toxicities. done by
network, Food Among other toxicities it applying several
and Drug predicts developmental QSAR
Administration, toxicity methodologies
and resulting in a
hierarchical greater
clustering confidence of
the results
TIMES QSAR statistical The TIMES platform is Classification Commercial
(COREPA) model based also used to predict models for the
Laboratory of different endpoints prediction of
Mathematical including receptor estrogen,
mediated endpoints for androgen and
 oestrogen, androgen and aryl
aryl hydrocarbon binding Hydrocarbon
affinity. predict binding. The
oestrogenicity chemical is
predicted to fall
in one of several
activity bins
(ranges of
binding affinity)
VirtualToxLab multi-dimensional predicting endocrine Classification Commercial
QSAR (mQSAR) disrupting potential by model for
simulating and endocrinedisrupt
quantifying the iong
interactions with aryl potential based
hydrocarbon, oestrogen on
alpha/beta, androgen, simulations of
thyroidalpha/beta, the interactions
glucocorticoid, liver X, towards
mineralocorticoid and aryl
peroxisome proliferator- hydrocarbon,
activated receptor estrogen,
gamma androgen,
thyroid,
glucocorticoid,
liver X,
mineralocorticoi
d, peroxisome
proliferator-
activated
receptor , as
well
as the enzymes
CYP450 3A4
and 2A13
OECD QSAR (Q)SAR Toolbox The Toolbox also Altough Freely
Application includes a range of primarily a tool Available
Toolbox profilers to quickly for chemical
evaluate chemicals for categories and
common mechanisms or read-acros, it
modes of action also includes
several
databases,
including
reprotoxicity
data: 166,072
ER binding data
from Danish
EPA (pre-
generated
predictions, not
 experimental
values) as well
as 1606
experimental ER
binding affinity
values from
the OASIS
commercial
database
Toxmatch 2D similarity Read-across predictions This study freely
methods were performed for the illustrates the available
17 query chemicals for use of 2D
which a category could similarity
be formed indices within
Toxmatch to
form categories
for 57 query
chemicals. The
underlying
hypothesis is
that chemicals
selected as being
similar should
act via a single
mechanism of
action, even if
that
mechanism is
unknown
CASE QSAR models to predict Ames test Improve current Commercial
(Computer mutagenicity using 520 predictive
Automated proprietary drug toxicology
Structure candidate (Test set 1) and capabilities for
Evaluation) 94 commercial (Test set mutagenicity
2) compounds and
carcinogenicity
through
customizing and
augmenting
current
predictive
software.
Prioritization of
synthesis &
testing
candidates.
Identification of
substructures
responsible for
an observed
 mutagenic
liability and
suggested
synthetic
alternatives.
Regulatory and
due diligence
support
Oncologic QSAR models Rule based expert system Knowledge Commercial
for the prediction of based expert
carcinogenicity system ,
mimicking the
decision logic of
human experts
Computer- Optimization of chemical Commercial
Aided 3D-receptor series (quality of leads)
Molecular modelling , All activities of assessment of
Modeling Virtual screening promising compound impurities,
(CAMM) , classes should focus on degradation
Fragment-based multiple ADMETox- products, side
screening related parameters in products,
parallel to activity and metabolites,...e.
selectivity Results of g. structural
commercially available evaluation of
tools for calculating synthesis
physicochemical schemes
properties and ADME-
related parameters have
to be interpreted with
great care, The use of
generic models can only
be recommended if they
have been validated for a
particular project; results
of new compounds
outside of the training
sets can be misleading
(ionization constants,
lipophilicity and
solubility) Shift in
optimization strategy, use
of measured values calls
for high quality, fast and
standardized assays
(100500 compounds per
week)
VEGA Caesar QSAR statistical The integrated model Quantitative AD Freely
model based utilizes two Measurement Available
on Neural complementary
Network + techniques in series: a
 rule-based model machine-learning
algorithm (support vector
machine), In the present
study, the conservative
approach was adopted,
and thus the chemicals
labelled as suspicious
were considered
mutagenic. The list of
structures, toxicity data,
and chemical
descriptors is available
online
LAZAR It is based on the toxicological endpoints performs Freely
use of statistical (currently, mutagenicity, automatic Available
algorithms for human liver toxicity, applicability
classification (k- rodent and hamster domain
nearest carcinogenicity 14 estimation
neighbours and and provides a
kernel models) confidence
and regression index for each
(multi-linear prediction, and
regression and is usable without
kernel models) expert
knowledge

Genotoxicity predicted system name, short description and predicted end points for

determination carcinogenicity and mutagenicity. (They are reported in following Table).

System name Short description Predicted end points
Classical QSAR Correlate structural or property descriptors of QSARs for various endpoints
approaches compounds with biological activities published
DEREK for Knowledge(rule)-based expert system M/C/SS/I and more (>40)
Windows
MCASE Machine-learning approach to identify Available modules:
(CASE, molecular fragments with a high M/C/T/I/H/MTD/BD/ATand
CASETOX) probability of being associated with an more
observed biological activity
Oncologic Knowledge-based expert system, mimicking C
the decision logic of human experts
MDL QSAR QSAR modeling system to establish structure- M/C/hERG inhib/AT/LD50
property relationships, create new calculators
and generate new compound libraries
Lazar Derives predictions from toxicity data by M/C/H/ET
searching the database for compounds that are
similar with respect to a given toxic activity
TOPKAT TOPKAT employs cross-validated QSTR Available modules:
models for assessing various measures of M/C/T/LD50/SS/I/ETand more
toxicity; each module consists of a specific
database
ToxScope ToxScope correlates toxicity information with M/C/I/H/T and more
structural features of chemical libraries, and
creates a data mining system
HazardExpert Knowledge(rule)-based expert system M/C/I/SS/IT/NT
COMPACT COMPACT is a procedure for the rapid C and P450-mediated toxicities
identification of potential carcinogenicity or
toxicities mediated by CYP450s
PASS Based on the comparison of new structures Multiple endpoints
with structures of well- known biological
activity profiles by using MNA structure
descriptors
 Cerius2 Molecular modeling software with a ADME/H
ADME/Toxtool package provides
computational models for the prediction of
ADME properties
Tox Boxes Modules generated by a machine-learning M/AT/C/LD50and more
approach implemented in a fragment-based
Advanced Algorithm Builder (AAB)
Meta Drug Assessment of toxicity by generating networks >40 QSAR models for
around proteins and genes ADME/Tox
(toxicogenomics platform) Properties
DICAS Cascade model with the capability to mine for C
local correlationsin datasets with large number
of attributes
C ADD Computer-aided drug design (CADD) by Receptor-and CYP450-mediated
multi-dimensional QSARs applied to toxicity- toxicities, ED
relevant targets
CSGenoTox QSTR-based package employing M
electrotopological state indexes, connectivity
indexes and shape indices
AdmensaInteractive QSAR-based system primarily for ADME CT
optimization
PreADMET Calculation of important descriptors and neural M/C
network for the construction of prediction
system
BfRDecision Rule-based system using physicochemical I and corrosion
Support System properties and substructures I

Contact:
Email: savalesagar484@gmail.com
Mob. No. 9960885333