chemistry and reactions from non- us patentsbulletin.acscinf.org/pdfs/247nm006.pdfus patents daniel...

248th ACS National Meeting, San Francisco CA, USA 10th August 2014

Chemistry and reactions from non-US patents

Daniel Lowe and Roger Sayle NextMove Software

Cambridge, UK


Topics

• Coverage of USPTO vs EPO patents

• Extraction of non-chemical-compound data

• Can text-mining provide insights into reaction informatics


What am I missing by only using the USPTO data?

Presenter

Presentation Notes

With Google Patents making the entire USPTO back-catalogue available for free using the USPTO data is an obvious starting point


EPO and USPTO background

• USPTO: 303k grants published in 2013

• EPO: 67k grants published in 2013

• EPO patents must be in one or more of English, French or German

• USPTO documents are all in English


Epo/USPTO Compounds (using StdInChI)

3,345,877

2,674,069

445,588

USPTO = 1976-2013 patent grants + 2001-2013 patent applications EPO = 1978-2013 patent grants and applications

Presenter

Presentation Notes

Lexichem was used to translate French and German text


Epo/USPTO Reactions (using separately sorted StdInChIs for reactants/agents and products)

706,524

317,533

99,681

USPTO = 1976-2013 patent grants + 2001-2013 patent applications EPO = 1978-2013 patent grants and applications

Presenter

Presentation Notes

No attempt at translating


Effect of different Languages

0

500000

1000000

1500000

2000000

2500000

3000000

3500000

4000000

1978

1979

1980

1981

1982

1983

1984

1985

1986

1987

1988

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

Num

ber o

f str

uctu

res e

xtra

cted

English

German

French

Translation performed by Lexichem v2014.Jun

Presenter

Presentation Notes

In the beginning English was far less dominant


First Disclosure Occurrence by Language

Excluding compounds disclosed by USPTO (1976-

2013) patents

All compounds

Presenter

Presentation Notes

Most compounds are first disclosed in English patents…. But most of these are also published by the USPTO! German and especially French patents less likely to be republished. Potentially warrants further investigation into French/German chemical name English translation software


Timeliness

• Competitive intelligence requires up to date information

• Methodology:

– Consider compounds present in both EPO and USPTO date not disclosed by either prior to 2006

– Compare the earliest publication date for each compound

Presenter

Presentation Notes

2006 chosen to rule out common chemicals and to eliminate the quirk that US patent applications are only available from mid 2001


0

20000

40000

60000

80000

100000

120000

140000

160000

180000

US morethan 5years

earlier

US 3-5years

earlier

US 2-3years

earlier

US 1-2years

earlier

US 6-12monthsearlier

US 3-6monthsearlier

US 1-3monthsearlier

US 1-4weeksearlier

Within aweek

US 1-4weekslater

US 1-3months

later

US 3-6months

later

US 6-12months

later

US 1-2yearslater

US 2-3yearslater

US 3-5yearslater

US morethan 5yearslater

Com

poun

d di

sclo

sure

s

Average lag between EPO/USPTO

Compounds first disclosed between 2006-2013

Mean: US 540 days earlier

Presenter

Presentation Notes

This isn’t the whole story. PCT applications are not republished by the EPO, hence often a US application is being compared to an EPO grant


0

10000

20000

30000

40000

50000

60000

70000

US morethan 5years

earlier

US 3-5years

earlier

US 2-3years

earlier

US 1-2years

earlier


US 3-6monthsearlier

US 1-3monthsearlier

US 1-4weeksearlier

Within aweek

US 1-4weekslater

US 1-3months

later

US 3-6months

later

US 6-12months

later

US 1-2yearslater

US 2-3yearslater

US 3-5yearslater


Com

poun

d di

sclo

sure

s

Applications Vs Applications




0

20000

40000

60000

80000

100000

120000

140000

US morethan 5years

earlier

US 3-5years

earlier

US 2-3years

earlier

US 1-2years

earlier


US 3-6monthsearlier

US 1-3monthsearlier

US 1-4weeksearlier

Within aweek

US 1-4weekslater

US 1-3months

later

US 3-6months

later

US 6-12months

later

US 1-2yearslater

US 2-3yearslater

US 3-5yearslater


Com

poun

d di

sclo

sure

s

Grants vs Grants




Presenter

Presentation Notes

As this is a PCT application, the EPO does not provide text for EP1005479A1. Using WO data would facilitate even more prompt discovery


Epo/US 1976-2013 and chembl19 overlap

187,486

1,155,034 6,278,048

Presenter

Presentation Notes

Similar analysis previously done vs SureChEMBL (http://www.chemaxon.com/wp-content/uploads/2014/03/EMBL-RBI-Mark-Davies.pdf) with a similar lack of overlap


0

1000

2000

3000

4000

5000

6000

7000

8000

9000

10000

Patents 4-5years earlier





Patents 0-1years later





Average lag between Patents/CHEMbl19


Mean: Patents 345 days earlier

Presenter

Presentation Notes

Not all records from ChEMBL are associated with years, degree of granularity in ChEMBL is only to the nearest year (approximated as the 1st of July for comparison)


Other data in patents

• Genes/proteins • Reactions including role of reagents • Physical quantities • Spectra • Diseases • Organisms • Companies • …

Presenter

Presentation Notes

What else can be extracted


Gene/protein identification

• Identify trends in drug target popularity

• Count number of patents (in a year) mentioning a given gene or its gene products and map to the HGNC symbol

• Many genes are referenced by short ambiguous terms hence great care is required to have good precision

Presenter

Presentation Notes

HGNC = HUGO Gene Nomenclature Committee



(Hepatocyte growth factor receptor) (Epidermal growth factor receptor)

0.00%

0.10%

0.20%

0.30%

0.40%

0.50%

0.60%

0.70%

0.80%

0.90%

0.00%

0.05%

0.10%

0.15%

0.20%

0.25%

0.30%

0.35%

1976197819801982198419861988199019921994199619982000200220042006200820102012

Perc

enta

ge o

f pro

tein

/gen

e m

entio

ns in

that

yea

r (Ch

EMBL

)

Perc

enta

ge o

f pro

tein

/gen

e m

entio

ns in

that

yea

r (Pa

tent

s)

EGFR

0.00%

0.10%

0.20%

0.30%

0.40%

0.50%

0.60%

0.70%

0.00%

0.02%

0.04%

0.06%

0.08%

0.10%

0.12%

0.14%

1976197819801982198419861988199019921994199619982000200220042006200820102012

Perc

enta

ge o

f pro

tein

/gen

e m

entio

ns in

that

yea

r (Ch

EMBL

)

Perc

enta

ge o

f pro

tein

/gen

e m

entio

ns in

that

yea

r (Pa

tent

s)

MET



(Mammalian target of rapamycin) (Tissue plasminogen activator)

0.00%

0.05%

0.10%

0.15%

0.20%

0.25%

0.30%

0.35%

0.40%

0.45%

0.00%

0.02%

0.04%

0.06%

0.08%

0.10%

0.12%

0.14%

1976197819801982198419861988199019921994199619982000200220042006200820102012

Perc

enta

ge o

f pro

tein

/gen

e m

entio

ns in

that

yea

r (Ch

EMBL

)

Perc

enta

ge o

f pro

tein

/gen

e m

entio

ns in

that

yea

r (Pa

tent

s)

MTOR

0.00%

0.05%

0.10%

0.15%

0.20%

0.25%

0.30%

0.35%

0.40%

0.00%

0.05%

0.10%

0.15%

0.20%

0.25%

0.30%

0.35%

0.40%

0.45%

0.50%

1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012

Perc

enta

ge o

f pro

tein

/gen

e m

entio

ns in

that

yea

r (Ch

EMBL

)

Perc

enta

ge o

f pro

tein

/gen

e m

entio

ns in

that

yea

r (Pa

tent

s)

PLAT

Presenter

Presentation Notes

MTOR rapid rise in application (same thing seen for grants, but with delay, ChEMBL also has a similar increase)


Presenter

Presentation Notes

Wordle based on gene/protein mentions in a major pharmaceutical’s 2013 US patent applications Limited to just mentions from titles/abstracts for improved precision. Can use analysis of previous years to identify new areas that a company is moving into


Melting/Boiling Point Extraction


Melting/Boiling Point Extraction

Over 99k so far!


CHEMICAL reactions

Presenter

Presentation Notes

A distinction between this resources and SCRIPDB is that it can be downloaded in its entirety for analysis


Predicting yield

• Features to consider: • 2D fingerprints (especially around the reaction

centers) • Reaction type • Temperature • Time • Change in complexity e.g. chiral centres


Data extraction

• Can pull out yields, quantities, times and temperatures

• Can sanity check text-mined yield by calculating from amounts (or even masses)

𝑚𝑚𝑚𝑚 𝑜𝑜 𝑐𝑜𝑚𝑐𝑜𝑐𝑐𝑐(𝑔)𝑚𝑜𝑚𝑚𝑚 𝑚𝑚𝑚𝑚 [𝑐𝑚𝑚𝑐 𝑜𝑚𝑜𝑚 𝑚𝑠𝑚𝑐𝑐𝑠𝑐𝑚𝑠](𝑔𝑚𝑜𝑚−1)

= 𝑚𝑜𝑚𝑚 𝑜𝑜 𝑐𝑜𝑚𝑐𝑜𝑐𝑐𝑐

𝑚𝑜𝑚𝑚 𝑜𝑜 𝑐𝑚𝑜𝑐𝑐𝑐𝑠𝑚𝑜𝑚𝑚 𝑜𝑜 𝑚𝑙𝑚𝑙𝑠𝑙𝑐𝑔 𝑚𝑠𝑚𝑐𝑠𝑚𝑐𝑠

= %𝑦𝑙𝑠𝑚𝑐

Presenter

Presentation Notes

Atom mapping used to determine which reactants contribute and hence which is limiting. Can still be tripped up by crude yield vs actual yield


Outliers inevitable


sCale vs yield

2001-2013 US applications, Suzuki couplings

Presenter

Presentation Notes

Same relationship found in GSK data by Nick Tomkinson (http://www.ukqsar.org/slides/5_Nick%20Tomkinson%20UKQSAR%2029th%20April%202014.pdf)


Temperatures

Presenter

Presentation Notes

Strong bias towards multiples of 10s (and to a lesser extend multiples of 5s). Notable exceptions at +-78oC Qualitative descriptions e.g. “Room temperature” not considered in this analysis


Identify Synthetic Routes

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17Intermediates 197702 103114 56611 31403 17268 9230 5057 2701 1256 639 301 136 58 15 5 2Terminal Products 385149 149445 81837 47579 27670 16619 9320 5263 2511 1330 678 373 111 63 8 6 5

0

100000

200000

300000

400000

500000

600000

700000

Occ

urre

nces

Number of steps

Intermediates

Terminal Products

Presenter

Presentation Notes

Steps to synthesise products in USPTO applications 2001-2013


Number of steps vs Complexity

ringComplexityScore = log10(nRingBridgeAtoms + 1) + log10(nSpiroAtoms + 1) stereoComplexityScore = log10(nStereoCenters + 1) macrocyclePenalty = log10(nMacrocycles + 1) sizePenalty = natoms1.005 − natoms

∑ Ertl & Schuffenhauer 2009 doi:10.1186/1758-2946-1-8

Presenter

Presentation Notes

USPTO applications 2001-2013 set


Number of steps vs Complexity

0

5

10

15

20

25

30

35

40

45

50

1 2 3 4 5 6 7 8 9 10 11

Num

ber o

f pro

duct

hea

vy a

tom

s

Number of steps

Average number of heavy atoms

Presenter

Presentation Notes

USPTO applications 2001-2013 set


Yield vs Change in Complexity

2001-2013 US applications, all reactions

Presenter

Presentation Notes

Reduction in complexity occurs with things like nitro reductions or deprotections Similar graphs with no clear trends can be generated for times/temperatures


Trends in Reaction Types

0.0%

1.0%

2.0%

3.0%

4.0%

5.0%

6.0%

7.0%

8.0%

1976

1977

1978

1979

1980

1981

1982

1983

1984

1985

1986

1987

1988

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

Suzu

ki c

oupl

ings

as a

per

cent

age

of re

actio

ns in

a y

ear

Classification performed using NameRXN

Presenter

Presentation Notes

USPTO grants 1976-2013 set


Trends In Solvent Use

0.0%

5.0%

10.0%

15.0%

20.0%

1976

1977

1978

1979

1980

1981

1982

1983

1984

1985

1986

1987

1988

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

Perc

enta

ge o

f rea

ctio

ns in

that

yea

r

Tetrahydrofuran

Dichloromethane

Water

Dimethylformamide

Methanol

Ethyl acetate

Ethanol

1,4-Dioxane

Toluene

Acetonitrile

Acetic acid

Chloroform

Acetone

Benzene

Presenter

Presentation Notes

Solvent usage of the top 10 solvents in 1976 and 2013 (USPTO grants 1976-2013 set)


Are solvents getting greener? 1976 2013

Water (21%) Tetrahydrofuran (15%)

Ethanol (11%) Dichloromethane (14%)

Benzene (8%) Water (13%)

Methanol (7%) Dimethylformamide (10%)

Tetrahydrofuran (5%) Methanol (8%)

Dichloromethane (4%) Ethyl acetate (7%)

Dimethylformamide (4%) Ethanol (5%)

Acetic acid (4%) 1,4-Dioxane (4%)

Chloroform (3%) Toluene (3%)

Acetone (3%) Acetonitrile (3%)

Total for top 10: 71% 82%

Presenter

Presentation Notes

Based on DOI: 10.1039/B711717E (Pfizer solvent selection guide) (USPTO grants 1976-2013 set)


Conclusions

• A significant amount of novel chemistry, from EPO patents, comes from the non-English patents

• Compounds disclosed by both the USPTO and EPO are on average published earlier by the USPTO but for many compounds an EPO patent will be the earlier disclosure

• Gene/protein identification can identify clear changes in patenting behaviour over time

• Text mining provides the tools to answer many reaction informatics questions


Acknowledgements

• Funding provided by:


Thank you for your time!

http://nextmovesoftware.com http://nextmovesoftware.com/blog

[email protected]

chemistry and reactions from non- us patentsbulletin.acscinf.org/pdfs/247nm006.pdfus patents daniel...

Documents