19
Ames/QSAR international collaborative project Masamitsu Honma, Ph.D. Division of Genetics and Mutagenesis, National Institute of Health Sciences, Kanagawa, JAPAN

Ames/QSAR international collaborative project

  • Upload
    others

  • View
    2

  • Download
    0

Embed Size (px)

Citation preview

Ames/QSAR international collaborative project

Masamitsu Honma, Ph.D.Division of Genetics and Mutagenesis,National Institute of Health Sciences,

Kanagawa, JAPAN

The Use of QSAR in ICH-M7 Guideline for Assessment of Genotoxic Impurities

Two (Q)SAR prediction methodologies that complement each other should be applied. One methodology should be expert rule based and the second methodology should be statistical based.

The absence of structural alerts from two complementary (Q)SAR methodologies (expert rule-based and statistical) is sufficient to conclude that the impurity is of no mutagenic concern, and no further testing is recommended.

How to Improve QSAR Prediction ?

New QSAR Algorithm/ Model• AI, Deep-learning ?

Training data set• New• Many• Reliable

Chemicals newly manufacturing produced or imported more than 100kg/year must be assessed its mutagenicity by Ames assay.

Industrial Safety and Health Law “An-eihou” in Japan

The permission of the use of the Ames data to improve QSAR models by Chemical Hazards Control Division, Industrial Safety and Health Department, Labor Standards Bureau in MHLW

0

5,000

10,000

15,000

20,000

25,000

30,000

1989

1990

1991

1992

1993

1994

1995

1996

1997

1998

1999

2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

2010

2011

2012

2013

2014

2015

Manufacture

import

Proposal of International Collaborative Studies to Improve

Ames/QSAR models (QSAR2014, Milan, Italy, June 2014)

1. Lhasa Limited (UK)

2. MultiCASE Inc (USA)

3. Leadscope Inc (USA)

4. Istituto di Ricerche Farmacologiche Mario Negiri (Italy)

5. LMC-Bourgas University (Bulgaria)

6. Istituto Superiore di Sanita (Italy)

4. Prous Institute (Spain)

5. Swedish Toxicology Science Research Center (Sweden)

9. FUJITSU KYUSHU SYSTEMS (Japan)

10. IdeaConsult Ltd. (Bulgaria)

11. Molecular Networks GmbH and Altamira LLC (USA)

12. Sumilation Plus (USA)

DEREK Nexus, SARAH

CASE Ultra rule-, statistical-based

Leadscope rule-, statistical-based

CAESER, SARPY, kNN

TIMES/AMES

Toxtree

Symmetry

AZAMES

ADMEWORKS

AMBIT

ChemiTunes and ToxGPS

Mut_Risk-0

QSAR Venders QSAR Model

Participants in Ames/QSAR Project

Ames Mutagenicity of Challenging Chemicals

Class A : Strongly positive, in which the chemical generally induces more than 1,000 colonies/mg in at least one Ames strains in the presence or absence of rat S9.

Class B : Positive, in which the chemical induces colonies more than 2-fold of the negative control at least one Ames strains in the presence or absence of rat S9, but not in class A.

Class C : Negative, which is neither class A nor B.

Category

Class A

Class B

Class C

Total

Phase I(2014-2015)

183 (4.7%)

383 (9.8%)

3,336 (85.5%)

3,902

Phase II(2015-2016)

253 (6.6%)

309 (8.1%)

3,267 (85.3%)

3,829

Phase III(2016-2017)

236 (5.4%)

393(8.9%)

3,780 (85.7%)

4,409

Total(2014-2017)

672 (5.5%)

1,085 (8.9%)

10,383 (85.6%)

12,140

Ames/QSAR Project (Phase I-III) Challenged Chemicals

100-Specificity (%)

Sens

itivi

ty (%

)

ROC Graphs for Challenged QSAR Models’ Validation

0

20

40

60

80

100

0

20

40

60

80

100

0 20 40 60 80 100 0 20 40 60 80 100 0 20 40 60 80 100

Validation I(A vs. C)

Validation II(A vs. C)

Validation III(A vs. C)

Validation I(AB vs. C)

Validation II(AB vs. C)

Validation III(AB vs. C)

CH3

N

H2N

Cl

Cl

O

Cl Cl

H

H3C

NH

O

O

O

CH3

OH

O

NH2

O

O

NH2OH

HO

H2N

HO

OH

NH2

CH

3

CH

3

CH

3

HC

CH3N

CH3

P

NH3CCH3

N

CH3

CH3

Cl

Cl

O

F F

F

FN

F

F

FCH3

P

Cl

Cl

H3CS

S NH

N

HNH

O

F F

F

NH

H

O

O

OCH3

OHO

CH3

O

OCH3

SO

O

Cl

H3C O

O

NH2 HCl

N

N

Cl

Cl

O

F

Cl

CH3

S

O

O Cl

CH3

CH3

Br

O

CH2

CH3H3C

CH3

False NegativesClass A chemicals, but negative call by almost QSAR tools

HH3C N

H

HH

O

OH3C

OF

F

FF

CH

F

FF

CH3

H3C

O

Cl

N

O

O

O O Br

Br

F

N

O

O

N N

O

FCl

O

O

N

F

F

O

O

F

O

N

O

ON

F

False PositiveClass C chemicals, but positive call by almost QSAR tools

O

N

O O

O

OO

O

H2N

O– N

+

O

O

O

N

N

NN

Cl

N +

OO–

Cl

H3C O

N+

O

O–

NH

O

N+

O

O–

O

N+

O

O–

O

N+

O

O–

O–

N+ O

O

O

N+

O–

O

HO

NN

HO

N+

O–

ON+

O–

O

O

NH

NH2

O

OH NOO

N

CH3

O O

NO

O

O

OCH3

H3C

N

NO O

N

N

CH3

N N

O

O

N

O

N

NH2

ClO

O

O

N

S

NO

O

O

N O

NO O

O

O

O

N

NO

OO

O O

O

O

N

NOO

O

O

O

O

OThey may be mutagens?

What means Ames positive?

Class A : Strongly positive, in which the chemical generally induces more than 1,000 colonies/mg in at least one Ames strains in the presence or absence of rat S9.

Class B : Positive, in which the chemical induces colonies more than 2-fold of the negative control at least one Ames strains in the presence or absence of rat S9, but not in class A.

Class C : Negative, which is neither class A nor B.

may contain false-positive.

may contain false-negative.

Is this Ames Positive?-Example A-

Confidential

Confidential

TA100 (+S9) 1st trial

Negative

TA100 (+S9) 2nd trial

Negative

TA98 (+S9) 1st trial

Negative

TA98 (+S9) 2nd trial

Negative

TA98 (+S9) 3rd trial

Negative

QSAR Tools Results

Derek NX INACTIVE

CASE UltraPHARM_SALM

Positive

QSAR Tools Results

Derek NX PLAUSIBLE

CASE UltraPHARM_SALM

Positive

4'-(chloroacetyl) acaetanilde

(Cas# 140-49-8)

O N

O

Cl

Is this Ames Positive?-Example B-

TA1537(-S9)

TA1537(+S9)

Lab. A Lab. B Lab. C Lab. D

Dunkel et al., Environ Mutagen, 7, Suppl. 5, 1-248 (1985)

Positive

QSAR Tools Results

Derek NX INACTIVE

CASE UltraPHARM_SALM

Negative

StructureAmesResult

Derek_NXver.4.0.5 Sarah

CASE UltraStatistical-

based

CASE UltraGT_EXPERT

Leadscope ICHM7 statistical-

based

Leadscope ICHM7 rule-based CAESAR SARPY TIMES/AMES Toxtree AZAMES ADMEWORKS AMBIT ChemiTunes MUT_Risk-0

C Posotive Positive Positive Positive Posotive others Positive Posotive Negative Positive Posotive N/A(1) Ambiguous Posotive Posotive

C Posotive Negative Positive Positive Posotive others Positive Posotive Posotive Positive Negative Posotive Negative Posotive other

C Posotive Positive Positive Positive Posotive others Positive Posotive Negative Positive unsure ? Positive Posotive Posotive

C Posotive Equivocal Positive Positive others others Positive Posotive Negative Positive positive ? Negative Posotive Posotive

C Posotive Negative Positive Positive Posotive others Positive Posotive positive Positive positive ? Positive Posotive Posotive

C Posotive Positive Positive Positive Posotive others Positive Posotive Negative Positive positive Posotive Ambiguous Posotive Posotive

HCl

NH 2

H 2 N N

ClO

FF

Cl O

FF

S

Cl

OO

CH3

O

Cl

O CH3

N

O

CH3

Cl

O

NH2

SO O

Cl

Cl

O

CH 3

Cl

O

Cl

O

N

NN

N

Org. Process Res. Dev., 2015, 19 (11), pp 1495–1506DOI: 10.1021/acs.oprd.5b00106

False Positive in Carboxylic and Sulfonic Acid Halides

Integrated approach for Genotoxicity Assessment

QSAR is not only a tool for the prediction. It can support to judge the results of actual Ames test.

Molecular Mechanism Mutagenic Potential

HO

HN

O

Cl

Cross-Talk

Re-modeling

Re-build Data Base

Feed-Back

QSAR beyond Ames

Summary

A large number of highly reliable data sets are essential to allow the development and improvement of QSAR models.

The Ames/QSAR international collaborative study is successfully ongoing. Its outcome gives a lot of benefits to QSAR vendors, QSAR users, and regulatory.

The integrated approach with QSAR results increases the sensitivity and specificity of the Ames test. It can support to judge the Ames results with molecular mechanism.

Web-Site of AMES/QSAR International Collaborative Study

AMES QSAR

The next Ames/QSAR challenge program will start from the end of 2018. Not only QSAR vendors, but also academia and IT companies are welcome to join the challenge. Hopefully, new QSAR models using AI and deep-learning will challenge.

AcknowledgementNational Institute of Health Sciences (Japan)

Airi KitazawaMasami YamadaTakeshi MoritaMakoto Hayashi

Lhasa Limited (UK)

MultiCASE Inc (USA)

Leadscope Inc (USA)

Prous Institute (Spain)

Bourgas University (Bulgaria)

Istituto Superiore di Sanita (Italy)

Istituto di Ricerche Farmacologiche Mario Negiri (Italy)

Swedish Toxicology Sciences Research Center (Sweden)

FUJITSU KYUSHU SYSTEMS (Japan)

IdeaConsult Ltd. (Bulgaria)

Molecular Networks GmbH and Altamira LLC (USA)

Sumilation Plus (USA)

Alex Cayley

Roustem Saiakhov

Glenn Matt

Christine DeMeo

Ovanes Mekenyan

Cecillia Bossa

Emilio Benfenati

Ulf Norinder

Hitomi Koga

Nina Jelazkova

Chihae Yang

Bob Clark

Chemical Hazards Control Division, Labor Standards Bureau in MHLW (Japan)

Kazuyo OofuchiShinji TsunodaHideaki HirakawaShinji KouzukiTatsuya Anai

Pharmaceutical Medical Devices Agency (Japan) Jun-ichi FukuchiKeiji Hirabayashi