A brief guide to the search tools and features in SureChem

You can use SureChem to search by chemical structure, chemical name, keyword or patent field. Structure searches are submitted using the Marvin drawing tool or entering a SMILES string. Searches begin with either a keyword or structure search, but once the initial result set is returned users can then filter using a combination of structures, keywords or patent fields. Search results, including chemical names and structures or patent text, can then be exported to the user's desktop if desired.

Structure Search

Select the 'Chemical Structure' search tab and then draw a structure or enter a SMILES string by copying and pasting into Marvin. Search options include similarity, sub-structure, full structure and duplicate match. Searches can be restricted to a particular field of a patent. For example by choosing 'Claims' only structures that appear in the claims section of a document will be returned.

Once a search has been completed the results are presented with a menu down the right edge of the screen to refine your search by changing the original query plus the ability to filter by a range of fields including keywords, assignees, inventors and date range.

The results can be displayed as a list of structures that match your search or as a list of patents that contain the structures that match your search. The patent listing display requires that the input structure returns less than 1000 similar or sub structures as the time required to search the millions of documents for over 1000 structures is prohibitive. If your input structure does return more than 1000 hits we recommend adding more detail to the structure and/or using the document field and keyword/assignee/inventor etc. filters to reduce the number of hits.

Basic Text Search

The Basic Text search tab allows a user to quickly perform a chemical name, SMILES, Standard InChI or InChIKey search in the 'Named Search' box or a keyword search in the 'Keyword Search' box.

The 'Named Search' automatically detects the type of query entered and then looks it up in the SureChem database. If it exists then the name it is mapped to and any synonyms which SureChem has linked to it are searched across the selected datasets. The search can be restricted to certain fields of a document using the 'Patent Field' selector.

For example if the Standard InChIKey 'FQCKMBLVYCEXJB-MNSAWQCASA-L' is entered as a search term it is first looked up in the SureChem database and is found to be the Standard InChIKey for the name 'Lipitor'. This name along with other synonyms that SureChem has mapped such as '134523-03-8', 'Atorvastatin', 'Atorvastatin calcium', 'CI 981', 'Xarator', etc. are then searched across the selected datasets.

The 'Keyword Search' box searches the selected datasets for the query term(s) as plain text. This search can also specify certain fields of a document using the 'Patent Field' selector.

The results of both the Named and Keyword searches are lists of patent documents that contain the query, and in the case of the Named search any synonyms of the query.

Advanced Text Search

Selecting the 'Advanced Text' search tab offers the ability to construct multiple term queries across a range of fields using the 'Boolean' and 'Complex search boxes.

Using the 'Boolean' search inputs you can enter keywords, chemical names, SMILES, InChI or InChIKey or any of the patent fields listed in drop-down menus if desired. If all the databases are selected then a generic list of patent fields covering all databases is displayed. Selecting a specific patent database will automatically bring up the patent fields specific to, for example, USPTO applications. Use the '+' or '-' buttons to add or delete keyword queries.

The 'Complex' search input area allows a user to construct a sophisticated query by directly specifying the query fields, boolean operators and lucene query syntax features to your search terms. Selecting a single database will give you access to fields specific to that data provider while selecting multiple databases will provide a set of field identifiers applicable to all.

Patent Number Search

Select the 'Patent Number' search tab and enter single or multiple patent numbers. SureChem recognizes up to 30 variants of patent number formats, so in most cases you should be able to enter the patent number as you found it, rather than trying to conform to a specific format, as is the case with other patent search tools.

Data Export (SureChem Pro only)

Once you have obtained search results, you can export them to your desktop using the 'Export' menu displayed on the left side of the screen. Chemical structure results include Patent ID, chemical name and SMILES strings, exported as an SDF or CSV file. Patent text can be exported as a Microsoft Word document or via email. To export the patent text including the marked-up chemical names, select the 'Print Screen' option and either print the file or save it as HTML.

Saved Search (SureChem Pro only)

Clicking on the 'Save Query' link at the top of the search results page will store your search criteria, sparing users the trouble of having to replicate complex but frequent searches. To access search queries you have saved, click on the 'Saved Searches' link in the upper right hand corner of the SureChem search page.

Alerts (SureChem Pro only)

SureChemPro users can get alerted when patents of interest are published. To set up an alert, click on the RSS icon , select your RSS reader and click 'Subscribe' (Firefox) or 'Subscribe to this feed' (Internet Explorer). Your search query as of that moment will be saved as an RSS feed. When new patents matching that query are published in SureChem an alert will be sent to your RSS reader. In Firefox you are also given the option of saving the feed as a live bookmark.

Frequently Asked Questions

1. What are the differences between full structure, duplicate, similarity and substructure search?
2. How can I refine my search results?
3. What does 'Complex Search' mean?
4. Do you get structures from images?
5. What does it mean when patents are ranked by 'confidence'?
6. SureChem seems to miss a lot of chemicals in the claims section. Why is that?
7. Does SureChem support Markush searching?
8. Can one configure RSS feeds for similarity search, or sub-structure, etc?
9. Are there any other options for exporting, such as adding titles, assignees, or where in the patent the compound was found?
10. Can I export a page showing all the chemicals highlighted?
11. Can I export multiple patents at once?
12. How can I see what chemicals SureChem has found in a patent?
13. Are you able to save searches?
14. Can you combine saved searches using Boolean operators? (e.g. Saved Search 1 and Saved Search 2 AND NOT Saved Search 3)?
15. Do you offer searches over a secure connection?
16. Do you offer any other ways of viewing the data?
17. Can you filter patents using IPC codes and other patent searching criteria?
18. What kind of search engine are you using for chemical search?
19. What kind of search engine are you using for text search?
20. How many chemical structures are in your database?
21. How many records are in your database?
22. What databases do you cover?
23. Will you be adding other patent databases?
24. How long does it take to update SureChem from when new patents are issued?
25. How is your synonym lists generated?

What are the differences between full structure, duplicate, similarity and substructure search?

All of SureChem's structure searches are powered by ChemAxon's JChemBase chemical structure search engine. For a full explanation of the different query types please refer to the ChemAxon search documentation at http://www.chemaxon.com/jchem/doc/user/query_searchtypes.html.

Full structure search attempts to bring back structures that are deemed to be the same as the input structure by taking into account special molecular properties. This search can return multiple hits

Duplicate search attempts to bring back a structure that is identical to the input structure.

Similarity search looks for structures that are either exact or close matches to your query structure. Two options for similarity threshold, 0.80 and 0.95 are currently offered. In future SureChem may offer a variety of similarity search algorithms.

Substructure search looks for structures that contain the query structure.

How can I refine my search results?

Structure, keyword and chemical searches can be "filtered" through a submenu on the right hand side of the results page. The refinement methods include date ranges, additional smiles similarity and substructure searching, inventor, assignee, and keyword. The returned results can also be refined to a single database if required.

Similarity and substructure search results can also be filtered by the different sections of a patent, e.g. selecting the 'Claims' option will limit the structures returned to those that occur in the Claims sections of the documents.

Boolean operators will also be added to the search filtering tool in future releases.

What does 'Complex Search' mean?

'Complex Search' provides the powerful command line search tools used in more traditional patent searches. Logical operators and Lucene search operators can be combined together to create complex queries that are not available in the graphical user search interface.

Do you get structures from images?

For the USPTO granted and USPTO Applications databases, structures are available for images of chemical compounds from 2001 onward. The images for a document are presented on a 'Chemical Image' tab in the document view.

What does it mean when patents are ranked by 'confidence'?

"Confidence" is a ranking based on the frequency of search terms within a patent. Those with a higher number of query terms have a higher confidence score. It is calculated by weighing the frequency of terms divided by the length of the document.

SureChem seems to miss a lot of chemicals in the claims section. Why is that?

As an automated system for extraction of chemical data from patent literature, SureChem is naturally stronger at detecting exemplified compounds and other whole compound names than generic chemical nomenclature terms such as (inorganic acid), fragments (nitro, cyano), classes (alkyloxy), and shorthand notation (CHCH3 group). We do recognize that some users are interested in searching these generic chemical terms and will be implementing means to do so in future SureChem releases.

Does SureChem support Markush searching?

SureChem currently supports basic radical and smarts queries. Future updates will expand this to more complex interactions using the Marvin drawing tool such as R-group and reaction queries. This will provide a framework for specific searches within classes of compounds and reaction types.

Can one configure RSS feeds for similarity search, or sub-structure, etc?

RSS feeds are available for any search where ever the "feed link" icon appears. Currently they are available only for patent search results within a single database using an exact match as the search method. The RSS feeds will be extended to cover the entire range of chemical and patent search options.

Are there any other options for exporting, such as adding titles, assignees, or where in the patent the compound was found?

In SureChemPro there are currently two forms of exporting enabled. The chemicals detected in the patent can be exported as SDF or CSV files for which the patent number is automatically added to the file name. Alternatively, or in conjunction, the text of the patent can be exported in a RTF format which contains the patent information such as the titles and assignees. There is no facility to mark up where the query term is located in the patent, nor is the document export a complete representation of the original patent.

Can I export a page showing all the chemicals highlighted?

The best way to capture the highlighting of chemicals is to copy/paste the contents of a patent to an HTML capable text editor, or to save the page as a complete web page using your browser. The markup will be intact with links through to SureChem for the highlighted chemicals. The markup will also print correctly.

The "printer-friendly" link reloads the current page without the advanced filtering and search queries enabling a better formatted document for physical printing. Using the "print preview" function (within the file menu of the web browser) allows the print format of the web page to be viewed and specific pages can be chosen for printing.

Can I export multiple patents at once?

Yes, from the search results page. By selecting the checkboxes next to individual results you can select a subset of the patents on the page to be exported. Due to the size and number of chemicals in some documents you are only able to select up to 3 patents when exporting all chemicals in a document. If your search returns less than 1000 documents you will be able to export the bibliographic content of all the results as Text or CSV.

How can I see what chemicals SureChem has found in a patent?

It depends on how you want to view them. Specific claims and examples are best viewed by the markup features while viewing the appropriate patent sections. To view the full set of chemicals detected within a patent, click on the "export" button and then open the results rather than saving them. The export contains the chemical names and smiles strings of all compounds converted.

Are you able to save searches?

In SureChemPro, searches can be saved, but must be used for single database searches. (In other words, you can't save a search across multiple patent databases). To save a search, click on the "SAVE" link to the right of the search query, and immediately after the RSS link.

Can you combine saved searches using Boolean operators? (e.g. Saved Search 1 and Saved Search 2 AND NOT Saved Search 3)?

No, not at this time.

Do you offer searches over a secure connection?

Yes, for SureChemPro users.

Do you offer any other ways of viewing the data?

Once a search is complete, the results can be displayed by Assignee and Inventor using the filtering tools.

Can you filter patents using IPC codes and other patent searching criteria?

This can be done using the drop down menu on the Boolean search page. Once the search is complete, filtering by IPC code is not possible.

What kind of search engine are you using for chemical search?

ChemAxon's JChemBase chemical structure search engine.

What kind of search engine are you using for text search?

Lucene, with additional chemical synonym associations developed by SureChem.

How many chemical structures are in your database?

Currently there are over 9 million unique structures.

How many records are in your database?

Approximately 14 million patents and 18 million MedLine abstracts.

For the coverage please see the data availability tab at the top of the page.

What databases do you cover?

Currently, the databases are:

• United States
  • Applications
  • Granted
• Europe
  • Applications
  • Granted
• WO/PCT
• Japanese Abstracts
• MedLine

Will you be adding other patent databases?

We are constantly looking to add more databases such as UK, Germany and Inpadoc. Users will be notified when they come online.

How long does it take to update SureChem from when new patents are issued?

From issuance to update on the website takes less than 24 hours. The exact dates and times depend on when the data is released which can vary with some of the issuing authorities.

How is your synonym lists generated?

We use a variety of sources to collate synonym lists including Open Access chemical databases and internal synonym matching. However, we are careful about using sources which have not gone through any type of manual curation as they often contain irrelevant synonyms for some compounds.

Summary:	Field  Value Total Patents 17,934,304 Total Unique Structures 11,323,904 Last Update Friday 5 March
Japanese Abstracts:	Field  From  To Documents – 8,125,842 Last Database Update – 2010-03-02 20:47:05 Issue Dates 1976-10-01 2009-11-26 Patent Number 1003204 2009278864  Chemical Information Unique Structures 242,876
MedLine:	Field  From  To Documents – 18,546,931 Last Database Update – 2009-12-11 06:40:00  Chemical Information Unique Structures 524,386
European Granted:	Field  From  To Documents – 859,870 Last Database Update – 2010-03-05 11:03:39 Issue Dates 1986-12-30 2010-03-03 Patent Number 256 2144227  Chemical Information Unique Structures 2,692,705
European Applications:	Field  From  To Documents – 1,382,455 Last Database Update – 2010-03-05 08:48:29 Issue Dates 1986-01-29 2010-03-03 Patent Number 169450 2160087  Chemical Information Unique Structures 3,032,116
USPTO Granted:	Field  From  To Documents – 4,123,217 Last Database Update – 2010-03-03 10:12:47 Issue Dates null 2010-03-02 Patent Number 3930271 7673347 Design Patents D242583 D611225 Defensive Patents T100001 T999003 Plant Patents PP3987 PP20809 Reissue Patents RE28671 RE41162 SIR Patents H1 H2234  Chemical Information Unique Structures 7,919,626
WO/PCT:	Field  From  To Documents – 1,339,656 Last Database Update – 2010-03-02 11:26:14 Issue Dates 1978-10-19 2010-02-25 Patent Number 1978000001 2010022410  Chemical Information Unique Structures 5,245,106
USPTO Applications:	Field  From  To Documents – 2,103,264 Last Database Update – 2010-03-04 23:18:19 Issue Dates 2001-03-15 2010-03-04 Patent Number 20010000001 20100058506  Chemical Information Unique Structures 6,084,770