• The search platform PubPharm is not a stand-alone subject database, but rather a discovery system, a modern detection system based on search engine technology. The data is organized into a single index to make it efficiently searchable. Discovery systems are part of the concept of library 2.0 and are intended to complement or even replace the previous Online Public Access Catalogs (OPACs).
• As of 01.10.2020 PubPharm offers approximately 55 million entries, of which over 30 million have been imported from the medical database Medline (PubMed). Additional resources originate from pharmaceutical, chemical and technological journals, preprints from bioRxiv, medRxiv, ChemRxiv, arXiv, engrXiv, Preprints.org and TechRxiv, books (e-books), dissertations, conference papers, information on clinical trials from the study registry ClinicalTrials.gov, and subject-specific patents from the European Patent Office.
• Pharmaceutical journals licensed (purchased) by the Specialised Information Service (SIS) Pharmacy are available to authorized users. Resources from Medline are updated daily, and from other sources at varying intervals from daily to monthly.
• The PubPharm search platform is freely accessible. Each user can perform searches and save the search results in a personal account.
• The availability of full texts differs according to user groups. Pharmaceutical scientists and other university staff members have full-text access to a wider range of journals than those interested outside. For more information, see How do I get the full text of an item?

The architecture of PubPharm

The PubPharm search platform is a discovery system. These systems are modern library catalogs based on search engine technology. Data from several databases are aggregated into a central search engine index. Among other sources, the entire Medline (PubMed), all holdings of literature on pharmacy and chemistry of the University Library Braunschweig and journals licensed by the SIS are available. All data are integrated into a single search index to make it searchable. With only one request journal articles (amongst others from Medline), textbooks or dissertations (from holdings of almost all university libraries in Germany as well as international dissertations) are being searched. Through this, the necessity for individually searching multiple databases is reduced.

The user interface (Frontend) of the search platform is based on the open source software VuFind and the beluga system of the Hamburg State and University Library (more information under https://qcovery.de, in german).

Full-text access

The search index contains metadata on each record. It provides information such as: title, author, publication year, format, etc. The list displayed as a result of your search is based on this metadata. Metadata usually does not contain the full text of a publication. For e-books and e-articles a link to the full text is usually indicated in the metadata or information is provided which can be used to generate such a link. If there is a link, it is displayed and you can directly access the publication. Access to the full text is dependent on several factors such as licence type.

It can happen that metadata does not contain a link to the full text. In these cases it is necessary to navigate via the journal homepage, volume and/or issue to get to the desired article.

Availability check

Depending on the user location, it is indicated whether a free full text access is available. The availability check is performed based on the user's IP address.

The availability check proceeds stepwise. In the initial step it is being determined whether the publication is freely available (Open Access). If it is a journal article, it is checked whether the article is licensed by the user´s local university library. A link resolver is used to examine the IP authorization for IP addresses belonging to german, austrian or swiss universities. To get access to full text versions it is important to be in your university network. If this is not the case at the moment, it is recommended to check whether your university offers a VPN service. In the next step it is checked if the item is licensed by the SIS (SIS-licences). After registration, authorized users have access to the full text version. In case there is no access to the full text version, a link to the commercial delivery system subito (www.subito.de) is offered. Alternatively, you can send a request to the SIS ( email to pubpharm [at] tu-braunschweig.de ). We will get back to you with information on how to obtain access to the full text. We cannot guarantee that we can deliver a copy of the full text immediately or directly.

• Standard search
  A standard query searches for the search term and similar terms with the same word stem (based on Porter Snowball).
• Search with wildcards (truncation or wildcards)
  When searching, a * can be used as a wildcard for any number of characters, anywhere or at the end of the word. It can not be used as the very first sign of a term. In principle, search is performed for search terms that share the same word beginning. For example, the output of a search for "hyperic*" contains results including the words hypericum, hypericin or hypericaceae. If ? is used as a wildcard at any position or at the end of the word, it stands for exactly one character. It is not possible to use ? at the beginning of the word.
• Exact word or phrase search using quotation marks "..."
  The phrase search "..." searches exactly for the specified word sequence. Wildcards / truncation (*,?) within quotes are not supported. In these cases, a search for a * or ? is conducted. Phrase search may be used for search terms that are located adjacent to each other ("capillary electrophoresis", "chiral separation"). This is also helpful if searching for authors "Steinhilber, Dieter" OR "Steinhilber, D".
• Use of word spacing operators/proximity search
  A word distance search can be triggered by adding a tilde (~) and a numeric value to the search phrase. For example, to search for "meta" and "analysis" within a maximum distance of two words between them, enter this search query: "meta analysis"~2.
•  

Combination of search terms (Boolean operators)

Boolean operators allow you to combine search terms. For example, their usage allows searching for titles that contain multiple search terms, titles that contain either one or the other keyword, or titles that contain a search term and do not contain another keyword. The use of parentheses makes it possible to refine search queries appropriately. The following is an explanation of the individual operators:

Using several search terms

• AND
  The AND operator is used by default. This means that the AND operation is used when no operator is placed between two words. When you combine words with AND, you obtain results that contain all the words. Example: You are looking for titles that contain the words "hot" and "melt" and "extrusion": hot melt extrusion is equivalent to hot AND melt AND extrusion.
• OR
  If there is an OR between two words, you get results in which one or both words were found. Example: You are looking for titles that contain the words aspirin or warfarin: aspirin OR warfarin.
• NOT or -
  Use the NOT operator to exclude results that contain this word. Example: You want to search for titles that contain the word hypericum but not the word perforatum: hypericum NOT perforatum, equivalent to hypericum -perforatum.

Usage of parentheses

• For more complex queries, using parentheses is useful for grouping search terms and Boolean operators. For example, you are looking for publications containing the phrases (see Exact word or phrase search) "vitamin c" or "ascorbic acid" and simultaneously either encapsulation, microencapsulation, or microsphere: ("vitamin c" OR "ascorbic acid") AND (encapsulation OR microencapsulation OR microsphere). It is important to use parentheses for such complex queries, and to insert an operator between all search terms. If NOT, - or OR are used somewhere in the search query, search terms between which there is no operator are combined with an OR during the search. Furthermore, the advanced search (see How can I use the advanced search?) can be used for complex queries. Requests are limited to a maximum of 500 characters in the search box.

• Change order of results
  By default, results are sorted by publication date (newest results first). Sorting by relevance (best matching article first), oldest results first, author (in alphabetically ascending order) or journal titles is also possible.
• Sort by relevance
  High weight is given to keywords, topics, name of the main author, title. Full text, publisher, ISBN, and ISSN have low weightings. The number of results displayed on a page is adjustable (10, 20, 50).
   
• Filter functions
  Results can be filtered by media type, journal title, year of publication, topic or language. Large result sets can be limited by filters; or you can continue the search retaining the currently set filters. For example, the user can limit the search to articles of a specific journal. Furthermore, yes-no-filters are offered for clinical trials from the study registry ClinicalTrials.gov, for patents and for systematic reviews. Through the filter "Systematic Reviews" results are retrieved, which have "systemat" and "review" in their title. This prototypical filter is currently undergoing improvements.

 All selected filters are marked with a checkmark and combined with a logical AND. Performed searches and filter settings can be recalled in the search history and saved in a personal account.

In the result list, you will find a summary of the most important information about each result.

You can see at a glance which resource it is (here: article), whether the full text is available directly (depending on the location), when the resource has been published and who the first three authors are. And if applicable, in which journal it has been published.

A new feature is the display of (possible) metrics on social media impact as published on Altmetric and (possible) citations of the resource, as far as they are stored in the database Dimensions. Data from both resources are available on PubPharm as far as there is information regarding the Digital Object Identifier (DOI) available and if the work has been cited or mentioned in the social networks at least once. For each new search, the citation/social media impact information is queried dynamically for providing the most current status. By clicking on the Dimensions icons, the user navigates to the corresponding page, where the citing articles/media are listed. Altmetric provides information on the social media response of the searched article. All contributions are individually clickable and can be studied in detail on the provider website.

In the detail view you will find detailed information about the respective result. If available in the index/dataset, you will find these details about a record:

• An abstract or breakdown into objectives, results, conclusions
• The list of all authors, where the first author is marked accordingly
• Under "Topics" the keywords for each result are listed
• Information on results such as ISSN / ISBN numbers, year of publication, DOI, source / copyright information

Depending on the resource, you will also find further information in the detail view:

Journal article

• Under "full text access" and "links" you may find several links to the full text. The location-dependent availability is marked with the respective tick (see How do I get the full text of an item?).

Journal title

• See "All Journal Articles" or "Associated Publications" and "more (+)" for all articles in the journal.

Book / E-Book

• For e-books, you will find several links to the full text under "full text access".
• For books and e-books that are not open access, there is no location-dependent availability check.
• Under "Links" there is often a link to the table of contents of the book.

Clinical Studies

• For each clinical study you can find the title, a brief study description if available, information on study phase and status, and identification numbers. Via the direct link (see "full text access") the complete entry can be opened on the website of the study register.

Patents

• For each result you can find information on patent classification (see "Keywords") and identification numbers (see "Patent ID"). Via the direct link (see "full text access") the complete entry can be opened on the website of the European Patent Office.

Why are there multiple links to the full text in the detail view?

In the detail view, all links are displayed that are stored in the data record for each result. The location-dependent availability of the individual links is marked with the appropriate check mark. Depending on which user group you belong to, different links can be interesting (for more information, see section How do I get the full text of an item?).

The result list contains icons that indicate whether the full text of a result is available directly (free of charge) or not. The availability check is location-dependent (based on the IP address of the search query). The full text is opened by clicking on the respective information text. The availability check enables the display of full-text access options at the respective location, and shows if licenses have been entered in the electronic journal database (EZB) or journal database (ZDB) for the respective institution. The local libraries are responsible for the timeliness of the EZB/ZDB entries. The majority of universities and publicly funded research institutions from Germany, Austria and Switzerland are listed in the EZB/ZDB. In addition, facilities from some other countries are also listed.

If possible, a direct link to the full text including information about the provider is shown. Sometimes it is not a direct link to the full text, but rather to the journal page. Here you can pick out the desired article on the basis of the publication year, issue, etc. In this case, the availability check also indicates whether the full text is freely accessible (). Not accessible full texts can be alternatively purchased using the commercial delivery system subito.

What can I do if I need the full text version of a journal article, but it is not directly accessible?

• If the article was purchased through the SIS licenses (symbolized by  ) and you are a researcher at a german pharmaceutical institute (authorized user group), you have free full text access after registration. It is required that your affiliation has obtained a licence and you use your university network when accessing PubPharm.
• If you have the opportunity, it is worthwhile dialing in via the IP network of a German university (e.g. via VPN service). The availability is checked depending on the location (based on the IP address). It could be, that the desired article has been licensed by your university library, and is therefore available to you.
• The detailed view of a result often contains several links. It is worth to check these. When checking availability, only certain links can be examined. However, it is possible that one of the other links provided offers free access to the full text.
• It may be worthwhile to search for the title of the desired article in another search engine, e.g. Google Scholar. Some full text versions of articles are accessible via institutional repositories or in social networks.
• The PubPharm team could help you when searching for print versions of articles listed by the PubPharm platform. In such cases, please use the Feedback Form or send your request to pubpharm [at] tu-braunschweig.de.

What can I do if I need the full text of a book (print or ebook) or a printed journal article?

• The availability check does not apply to printed books. For e-books, only open access availability is checked. That is, it is worth looking for the book you want in the local library catalog. If the desired book is not there, it can be ordered by the local library as an inter-library loan. Alternatively, the full text can be purchased via the commercial delivery system subito, if available.

Favourites lists

• By clicking on the favourites button (star), individual results or all results on a given page below Filter & Sort Options can be added to the favourites list. From this it is possible to export all or selected titles, to send them as a link list by email to print or to save them in a permanent list in the user account.   
• You can create favourites lists in a browser session without logging in (the favourites will be emptied after closing the browser), or you can create a free account and compose personal lists, which you can then edit at the next login.
• Selected items can be sent via email, printed or saved to the personal account, or exported to reference management programs.
   

Personal account

• Everyone may create their own free personal account to save favourites lists and search queries.
• You can create your personal account in the menu Login and Create New Account.
• You can save an item in a new or an existing list by clicking on "Save".
• Links to the favourite list can be sent by email; individual elements can be deleted or exported to a reference management program. Sorting by relevance, oldest results first, author or journal titles is also possible for items in the favourite list.
   

What can the export function for reference management software be used for?

• A reference management tool or reference manager contains a database, into which data about publications can be imported and managed. The linkage with a word processing program makes it possible to automatically insert references with a list of references in a specific citation style.
• Results in PubPharm can be directly imported into a reference manager via a favourite list. The following export formats are provided: BibTex, EndNote, EndNoteWeb, RefWorks, MARC, MARCXML, RDF, and CSV. In addition, corresponding IDs for each result, such as DOIs, if available, and COinS in the interface, are offered in such a way that browser extensions of reference management tools, such as Citavi or Zotero, recognize them and enable their further use. Well known reference management programs are e.g. Citavi, EndNote, Mendeley or Zotero.

How can I be notified automatically about newly indexed results?

• An RSS feed for a search can be subscribed. This shows newly added entries with title and year of publication. Once the user has used the search box, the RSS feed appears at the bottom of the results page. Feed-Reader offer the possibilty to manage and to read a RSS-Feed, and to inform you if new results are available.

The Advanced search supports formulating more complex queries. The link to the advanced search can be found below the search box.

In the advanced search, you can set which index fields (areas in the search index) should be used for which search term. For example, you can search for all the publications of an author in a particular journal: The following example illustrates how to search for publications by Prof. Stefan Laufer in the journal "Nature Medicine".

These index fields can be selected in the advanced search:

• Standard search: Search in fields with main contents: Author(s), title, keywords, abstract, table of contents, publisher, ISBN/ISSN.
• Title: Search within the title (e.g. titles of articles and books).
• Journal Title: Search within the title of journals. Also articles published in the particular journal will be returned.
• Abstract: Pre-set as part of the standard search or via Index (All Fields) in the advanced search.
• Person: Search for authors (including co-authors and other contributors).
• Publisher: Search for the publisher (for books and journals).
• Subject: It searches for publications on a specific topic. Each article in PubPharm is automatically assigned to specific topics that result from phrases and keywords. These words are available as part of the metadata for the selected article. In most cases, vocabulary words are used (such as Medical Subject Headings, MeSH). All data records from the Medline in the search index should be provided with corresponding entries from medical subject headings. Other records may also contain both controlled vocabulary entries and free keywords.
• Full text: The search term is searched in the full text. It should be noted that full texts are not available in the database for all publications.
• Table of contents: Search within the table of contents.
• ISBN/ISSN: Search for the ISBN of books or ISSN of journals.
• Publication year: Search for the publication year.
• Index (all fields): In addition to the standard search, additional fields, such as URLs, are parsed. These search functions provide the most comprehensive search.

Search Fields

• AND Returns only records that match every search field.
• NOT Returns all records EXCEPT those that match search fields.

The "Add Search Field" button may be used to add additional search fields to the form. You may use as many search fields as you wish.

Search Groups

For certain complex searches, a single set of search fields may not be enough. Search Groups provide a way to build searches from multiple groups of search fields. Every time you click the "Add Search Group" button, a new group of fields is added. Once you have multiple search groups, you can remove unwanted groups with the "Remove Search Group" button, and you can specify whether you want to combine search groups by AND or OR.

Example
For example, suppose you want to find results dealing with Bisoprolol and Captopril for the treatment of hypertension. In the first search group, enter "Bisoprolol" and "hypertension" and make sure that the "Search Operator" group setting is "AND". Add a second search group and enter "Captopril" and "hypertension". Make sure the search operator next to the Search Groups header is set to "OR".

The structure search is based on data from the database PubChem, further information can be found in this section under "Background information on structure search"

Apart from the text/advanced search options provided with the search box, the structure search allows searching for chemical compounds by entering their molecular structure, name, or import of SMILES-/InChI Code. Furthermore, once the substance has been identified, substructure and similarity search options are available. Here, further information can be obtained by clicking on the links leading to the PubChem and DrugBank database webpages. Publications related to an identified substance can be searched within PubPharm. Additionally, further substances that are structurally similar to the initial chemical compound can be determined.

Starting the search

There are the following search options:

• Draw the compound in the editor and run a search by clicking on: . Information regarding the editor can be found on the help pages of the Epam Ketcher.
• Enter the name of a substance and search for it by clicking on: . The detected compound is loaded into the editor where it can be modified.
• Insert the InChI or SMILES code of a chemical compound and click on  to load it into the editor. The structure can be modified in the editor; compounds can be searched for by clicking on:  .

Result list

If a compound is found in the database, a short overview (PubChem) is displayed. If available, a short description from DrugBank is shown as well. Via  it is possible to search directly in PubPharm for publications regarding the particular compound. In such cases, the first 5 compound synonyms retrieved from PubChem are used for the search.

In addition to brief information regarding the compound, links to corresponding entries in PubChem, DrugBank and the ChEMBL database (if available) are displayed.
The link  can be used to search for patents, in which the particular compound has been mentioned. It leads to the search interface of the European Patent Office (EPO).

Further search options

Starting from a compound in a result list, the structure search can be continued:

• Substructure search: Searches for compounds that contain the substance as a substructure. 
• Similarity search: Searches for compounds that are similar in structure. 

A list of the search results is displayed. Each result can again be used as a starting compound for a new search. Via  the structure can be loaded into the editor where it can be modified.

If a structure was entered in the editor for which no proper corresponding substance could be found in the PubChem database, the result gets marked with . It is then still possible to perform a structure or similarity search based on the compound.

Background information on structure search

The structure search is based on data from the PubChem database. Via the PUG REST API, search results are determined and processed for display. The compound search (based on either the structure entered into the editor or the name search) as well as the substructure and similarity search are conducted via the PubChem PUG REST API. Thus, search parameters used by PubChem and fingerprints, where necessary, are applied.

The innovative search tools, which are unique features for PubPharm, were and are actively developed in close cooperation with the Institute for Information Systems (IfIS) of the Technische Universität Braunschweig. For example, a semantically enriched search based on machine-learning technology offers a completely new PubPharm search feature to make searching for and accessing relevant information easier, more efficient and more effective. One example is the context-based faceting of drugs. Here, text-mining processes link bioactive compounds with contextually similar diseases and genes.

Based on the IfIS research a webservice was already implemented in PubPharm: when searching for a drug substance or a disease/symptom, for example the inhibitor of dipeptidyl peptidase-4 sitagliptin, a suggestion list of related, contextually similar substances, diseases / symptoms and genes is generated. Further information can be obtained via links to the Braunschweig Enzyme Database, BRENDA, and to the databases UniProt and KEGG.

In addition to sitagliptin, other drugs, such as alogliptin and exenatide, are listed among the related substances used to treat diabetes mellitus. Among the related diseases in this example, symptoms such as hyperglycemia and hypoglycemia are suggested. Another example of generating related substances is the listing of kinase inhibitors when entering the compound name vemurafenib. An example search with the active ingredient sumatriptan can be found here, and with the search term "arthritis" here .

• The website of the PubPharm search platform is responsive. This means that the website is displayed well on various devices - from the smartphone to the desktop PC.
• No matter which device you are looking for, the same database is searched and you can use all functions, such as filters, favourite lists and exports. The only exception is the structure search component.
• It should be noted that the layout arrangement of the individual elements is partially different. For example, the desktop PC view filter options are located in a column next to the result column, while in the smartphone view they are located under the result list.

The PubPharm project team at the University Library Braunschweig welcomes questions and suggestions regarding the PubPharm research platform. You can contact us as follows:

• By the Contact page
• By email: pubpharm [at] tu-braunschweig.de
• By phone: 0531 / 391 5046 or -5003
• By letter or in person: Fachinformationsdienst Pharmazie, Braunschweig University Library, Universitätsplatz 1, 38106 Braunschweig

The PubPharm research platform is continuously developed and optimized. Comments, suggestions for improvement or other feedback are always welcome!

The PubPharm Blog, our Twitter, Mastodon and LinkedIn accounts provide information on current developments and various work areas of the FID Pharmacy.