Transcript 13Segment - Ganesha Associates

Basic reading, writing and
informatics skills for biomedical
research
Segment 3. Using the virtual
library
10 April 2008
Copyright: Ganesha Associates
2008
1
The scientific research process
Experiments
Problem definition
Grant proposal
More
experiments
Seminar
Conference proceedings
Writing-up
Publication
Key:
Experiments
Reading and writing
Verbal communication
10 April 2008
Copyright: Ganesha Associates
2008
2
What is a literature review ?
• A literature review is a survey of what has been published in peerreviewed scientific journals about a particular research question at
this moment in time
• The aim is to provide the reader with an overview of the knowledge
theories and main areas of discourse within a particular subject
• It will include descriptions and summaries of key ideas, plus critical
evaluation of the ideas being presented, comparison between
commentators etc.
• As a piece of writing, the literature review must be defined by a
guiding concept - your research objective
• So it is not just a descriptive list of the material available, or a set
of summaries – it is a proposal, a plan of action, based on a solid
foundation of evidence
10 April 2008
Copyright: Ganesha Associates
2008
3
What is a literature review ?
• Start from a general perspective to orient the
reader to the importance of field to are working in
• But bring them quickly to the areas that of
particular importance today
• Provide a logical framework for selecting one of
these problems
• And then explain how the problem can be solved,
giving the reasons for the approach you have
selected
10 April 2008
Copyright: Ganesha Associates
2008
4
Finding a place to start
• Find the frontier
–
–
–
–
Scientific weeklies, Nature, Science
Top international journals in your field
Conference reports
Talk with a leader in the field
• Understand the background
– Review journals
• Trends in…, Current Opinions in…, Annual Reviews in…
– Use ‘cited by’ functionality in Scopus and WoS
– Use ‘related article’ functionality in PubMed
10 April 2008
Copyright: Ganesha Associates
2008
5
“Little is known about…”
• Means either the area is of little interest, or
you haven’t done your literature search
effectively
• Ask an expert
• Use a comparative or inductive approach
and look for examples in similar systems this technique is particularly useful in any
field with a molecular component
10 April 2008
Copyright: Ganesha Associates
2008
6
Unfortunately today search =
document retrieval
• Get me the abstract/full text for this
specific citation
• What is the latest research on subject x ?
• What is the latest review on subject y ?
• Who has cited this article recently
• What labs are working on this problem ?
10 April 2008
Copyright: Ganesha Associates
2008
7
How many piano tuners live in Chicago ?
• What genes are involved in the melanogenesis
of human lung cancers?
• What proteins in rats perform functions different
from those of their human homologs?
• What is the drug of choice for condition x ?
• Can drug x cause (adverse) finding y ?
• Find safety-related information associated with
GSK product x, competitor product y and drug
class z
10 April 2008
Copyright: Ganesha Associates
2008
8
Can drug x cause (adverse) finding y ?
• Drug x synonyms, spelling ?
• Finding y synonyms
• “Adverse”, “side effect”, etc
• Query finds all documents containing
these words, regardless of causal
relationship
10 April 2008
Copyright: Ganesha Associates
2008
9
How do search engines find things ?
• All documents are indexed by building a list of
key words, often called an index
• The index algorithm will make an entry for
each term or word found in a document
• Stop words, such as "the", “by”, “within” which
are both too common and carry too little
meaning to be useful for searching are
ignored
10 April 2008
Copyright: Ganesha Associates
2008
10
The structure of an A&I database
record
• Journal name, ISSN, date of publication,
volume, page numbers
• Authors names
• Address, affiliation, contact details
• Titles
• Keywords
• Abstract
• Metadata
10 April 2008
Copyright: Ganesha Associates
2008
11
How do search engines find things ?
• Some indexers also use ‘stemming’, for
example any of the words “enzyme",
“enzymatically", or “enzymatic" will be
recorded in the index under “enzyme"
• At the retrieval stage only the index is
searched rather than the text of the original
documents
• Documents containing the search terms can
be ranked in a variety of very different ways
10 April 2008
Copyright: Ganesha Associates
2008
12
How do search engines find things ?
• PubMed ranks by data of acquisition
• Scopus and Web of Science rank either by
data of publication or relevance
• Relevance is calculated by a formula which
takes into account of where and how often the
search terms appear in the document
• Even if the search engines searched the same
content (they don’t), they would order the
results differently
• Most people don’t look beyond the first page
10 April 2008
Copyright: Ganesha Associates
2008
13
How do humans find things ?
• They don’t look beyond the first page
• They select articles by title or (less often)
author, rarely by journal
• They often make little attempt to modify their
search terms if they cannot find what they
want
10 April 2008
Copyright: Ganesha Associates
2008
14
So, use keywords carefully
Question 1: What is the optimal management strategy for radial head
fractures based on randomized controlled trials?
Keywords: radial head fractures randomized
Result: 9 records found, one judged highly relevant
Question 2: Paper by Glasziou on radial fractures in the BMJ in 2007?
Keywords: glasziou fractures bmj 2007
Result: 1 record (the target) found.
Question 3: Review article on the state of vitreous body (of the eye)
and time of death?
Keywords: vitreous body time death
Result: 7 records found, several relevant.
Note: These examples are from PubMed
10 April 2008
Copyright: Ganesha Associates
2008
15
Improving recall precision
Keywords. Trained indexers supply a list of words
and their synonyms that describe the subject of
the text, e.g. MeSH
Field-restricted search. Some search engines
enable users to limit free text searches to a
particular field within a stored data record, such
as "Title" or "Author.“
Boolean queries. Searches that use Boolean
operators (for example, “drug" AND “side effect"
NOT “headache") can dramatically increase the
precision of a free text search.
10 April 2008
Copyright: Ganesha Associates
2008
16
Improving recall precision
Phrase search. A phrase search matches only
those documents that contain a specified
phrase, such as “cancerous growth"
Proximity search. A phrase search matches only
those documents that contain two or more words
that are separated by a specified number of
words
Wildcard search. A search that substitutes one or
more characters in a search query for a wildcard
character such as an asterisk. For example,
"s*n" will find "sin", "son", "sun", etc. in a text.
10 April 2008
Copyright: Ganesha Associates
2008
17
Real queries - Google
•
•
•
•
•
•
•
pharmacogenomics and disorders
bacteria growth casein media effect
waal pseudomonas
TRPM2 PCR mouse
Chitinases in carnivorous plants
glycerophosphoinositol 4-phosphate
Dai N, Gubler C, Hengstler P, Meyenberger C,
Bauerfeind P. Improved Capsule Endoscopy
After Bowel Preparation. Gastrointest Endosc
2005;61(1) 28-31.
10 April 2008
Copyright: Ganesha Associates
2008
18
Real queries - PubMed
• ATR1 HAL2
• Fuzzy[ALL] AND Hanage[AU] AND
2005[DP]
• arndt and rhabdomyosarcoma
• "Vorster HH"[Author]
• (rotavirus infections[majr] OR
rotavirus[majr]) AND english[la] AND
humans[mh] NOT (editorial[pt] OR
letter[pt])
10 April 2008
Copyright: Ganesha Associates
2008
19
More examples of search terms PubMed
hurley M J [auth]
AK230159
cytokinin signalling in arabidopsis
A genetic screen implicates miRNA-372 and miRNA-373 as
oncogenes in testicular germ cell tumors
cytochrome P450cam gene sequence
target of wingless in drosophila
10 April 2008
Copyright: Ganesha Associates
2008
20
Search terms – PubMed via Google
Diabetes high blood pressure losartan potassium
ET-1 AND sepsis AND heart
17[volume] AND 17[page] AND 2004[pdat]
beta-galactosidase + plant + flower
"transient K+ current" motoneuron
determining time of death by body temperature
10 April 2008
Copyright: Ganesha Associates
2008
21
Query changes people actually
make
• Query 1
–
–
–
–
–
latrunculin
latrunculin fm3a cell arrest
latrunculin fm3a arrest
latrunculin fm3a
latrunculin FM3A
• Query 2
–
–
–
–
cytokinin signalling in arabidopsis
"cytokinin signalling in arabidopsis"
cytokinin delta
spindly arabidopsis
• Results
– Remember to look beyond the first page
10 April 2008
Copyright: Ganesha Associates
2008
22
Search terms - summary
• Make sure you understand the search term syntax used
by your preferred site, i.e. AND, +, “, etc
• Search engines ‘see’ only certain words, not sentences
• Do not use ‘stop’ words, i.e. a, the, of, before unless
they are part of “a text string search”
• Google can be an effective tool for searching Entrez
databases
10 April 2008
Copyright: Ganesha Associates
2008
23
Special features: Google
• Unlike PubMed, Google searches web pages.
• Google supports OR and NOT Boolean operators.
• Submitting a query to Google in the form [propanolol
hypertension] will return all the web pages that match
both "propanolol" and "hypertension" exactly
• However, some of the terms may reside not on the
retrieved page itself, but on pages that link to the
retrieved page
• Instead of returning web pages chronologically, Google
employs a unique method called PageRank and
sophisticated text-matching techniques to find pages that
are both important and relevant to a query.
10 April 2008
Copyright: Ganesha Associates
2008
24
Special features : PubMed
• PubMed is a service of the National Library of
Medicine that includes over 15 million
bibliographic citations from MEDLINE and other
life science journals for biomedical articles back
to the 1950s.
• The full text of articles is not stored by the
system – only “header” information (title, authors
names, abstract, etc)
• Each article is indexed according to multiple
fields, including title, abstract, author names,
journal name, language of publication, year of
publication, etc.
10 April 2008
Copyright: Ganesha Associates
2008
25
Special features : PubMed
• Each article in MEDLINE is also indexed using a
controlled vocabulary, called Medical Subject
Headings (MeSH), which is used to describe the
main topics discussed.
• The set of MeSH terms is manually assigned by
biomedical experts who scan each article.
• Results are ranked in reverse chronological
order (default setting)
10 April 2008
Copyright: Ganesha Associates
2008
26
Special features : Scopus
• Scopus is the largest abstract and citation database. It
covers:
• 15,000 peer-reviewed journals from more than 4,000
international publishers, including coverage of:
–
–
–
–
Over 1200 Open Access journals
500 Conference Proceedings
Over 600 Trade Publications
200 book series
• 33 million records, of which:
– 16 million records include references going back to 1996 (75%
include references)
– 17 million pre-1996 records go back as far as 1841
• Scopus also covers 386 million quality web sources,
including 22 million patents.
10 April 2008
Copyright: Ganesha Associates
2008
27
Special features : ScienceDirect
• ScienceDirect (Elsevier) contains over 25% of the
world's science, technology and medicine full text and
bibliographic information.
• Apart from online eBooks, Reference Works, Handbooks
and Book Series ScienceDirect offers a rich journal
collection of over 2,500 titles.
• In addition, the Backfiles program offers the ability to
search a historical archive of over 6.75 million articles
directly from your desktop, back to Volume 1, Issue 1.
The collections contain four million articles prior to 1995,
and 2.75 million articles from after 1994.
10 April 2008
Copyright: Ganesha Associates
2008
28
Special features : Web of Science
• The Web of Science provides seamless access to
current and retrospective multidisciplinary information
from approximately 8,700 of the most prestigious, high
impact research journals in the world.
• Web of Science also provides a unique search method,
cited reference searching.
• With it, users can navigate forward, backward, and
through the literature, searching all disciplines and time
spans to uncover all the information relevant to their
research.
• Users can also navigate to electronic full-text journal
articles.
10 April 2008
Copyright: Ganesha Associates
2008
29
Search engine ‘arms race’
• Scopus and Web of Science will continue to
compete to provide user friendly tools for
document retrieval
• Their coverage will gradually expand to include
all jouenals (including national ones)
• PubMed (and NCBI) will focus on making it
easier to search across different types of
biomedical data
• Journal coverage will remain limited compared
to Scopus and Web of Science
10 April 2008
Copyright: Ganesha Associates
2008
30
Additional features
• Related articles
– PubMed
• Cited by
– Scopus, WoS
• Curated links to other forms of related
material
– PubMed
• Personalisation features e.g. MyNCBI
– PubMed, Scopus, WoS
10 April 2008
Copyright: Ganesha Associates
2008
31
Set up alerts
• Search alerts send new results from a
previous search directly to your mailbox.
• Document citation alerts notify you when
an article you specify is cited by another
article.
• Author citation alerts to notify you when
any of an author's publications get cited.
• The alerts can run daily, weekly or monthly
and deliver new results to you by e-mail.
10 April 2008
Copyright: Ganesha Associates
2008
32
About RSS
• RSS (Really Simple Syndication) is an XML-based
syndication format for distributing content on the Web.
Many kinds of content can be distributed using the RSS
format including Scopus® search and citation results.
• Scopus offers XML-based feeds so you can receive
updates on your desktop of all newly loaded content
matching your search criteria. Each feed contains a list
of the most recent titles which link back to the
appropriate article in Scopus. Scopus content is
refreshed every day.
• Scopus feeds are in RSS 2.0 format. To use a Scopus
RSS feed, you must have an RSS reader.
10 April 2008
Copyright: Ganesha Associates
2008
33
Cited by…
10 April 2008
Copyright: Ganesha Associates
2008
34
Why things sometimes don’t work
10 April 2008
Copyright: Ganesha Associates
2008
35
Managing your literature research
on your computer
• Download PDF/HTML versions of
important documents
• Give the files concise explanatory titles so
you will remember what they are about
• Ditto folder names
• Download Google Desktop Search, a
utility that indexes all the files on your PC
and displays results in a Google search
page
10 April 2008
Copyright: Ganesha Associates
2008
36
Citation management software
• Zotero is a free, easy-to-use Firefox extension that helps
you collect, manage, and cite your research sources
• Use EndNote to search online bibliographic databases,
organize their references, images and PDFs in any
language, and create bibliographies and figure lists
• Reference Manager is the industry standard software
tools for publishing and managing bibliographies on the
Windows and Macintosh desktop
• Confused ? Go to
http://en.wikipedia.org/wiki/Comparison_of_reference_m
anagement_software for a comparison of the features of
these and other software packages
10 April 2008
Copyright: Ganesha Associates
2008
37
Starting to search
10 April 2008
Copyright: Ganesha Associates
2008
38
Initial choices
?
Scielo
CAPES
Portal
National OA: BMC
Literature Or PLoS
10 April 2008
Web of
Science
Scopus
Science
Direct
Springer
Link
PubMed
Google
Other
HighWire Databases,
e.g. NCBI
Copyright: Ganesha Associates
2008
39
Anatomy of a session
GOOGLE search
Search results
10 April 2008
GOOGLE search
AbstractPlus
AbstractPlus
Full Text
Search results
BLAST results
Copyright: Ganesha Associates
2008
AbstractPlus
Full Text
NCBI full_report
Full Text
40
Google search
10 April 2008
Copyright: Ganesha Associates
2008
41
PubMed search
10 April 2008
Copyright: Ganesha Associates
2008
42
Scopus search
10 April 2008
Copyright: Ganesha Associates
2008
43
Access to full-text depends on
whether…
The journal is on CAPES list
CAPES has a current subscription with the publisher
The publisher has provided linking information to PubMed/Scopus, etc
The link points to a source which recognises your institution as a CAPES
subscriber
The journal has an open access policy after 0, 6, 12, 24 months
Many journals are available from more than one site, but not all may
recognise you as a subscriber
Remember: Subscribed-to content is only available on campus
10 April 2008
Copyright: Ganesha Associates
2008
44
PubMed search
10 April 2008
Copyright: Ganesha Associates
2008
45
PubMed Search
10 April 2008
Copyright: Ganesha Associates
2008
46
PubMed Search
10 April 2008
Copyright: Ganesha Associates
2008
47
Scopus search
10 April 2008
Copyright: Ganesha Associates
2008
48
Scopus search
10 April 2008
Copyright: Ganesha Associates
2008
49
Scopus search
10 April 2008
Copyright: Ganesha Associates
2008
50
ISI Web of Science
10 April 2008
Copyright: Ganesha Associates
2008
51
Why should I change my strategy?
Too Many results:
• Choose additional limits or add terms to
your search set
• Use the search history option to combine
different terms together
• Look at the index terms common to “good
articles”
• Check your Boolean operators; consider
using and or not
10 April 2008
Copyright: Ganesha Associates
2008
52
Why should I change my strategy?
Not Enough Results:
• Consider alternate or related keywords for
your topic
• Search abstract, title, headings for words
representing key concepts
• Choose keyword search option in
MEDLINE for recent topic/unique phrase
• Consider what title of ideal article would be
and use these terms
10 April 2008
Copyright: Ganesha Associates
2008
53
Why should I change my strategy?
Few Relevant Results:
• Use the related articles features in
MEDLINE, Scopus and Web of Knowledge
to match articles with a relevant result
• Enter the information from a good article in
Scopus or Web of Knowledge to track
articles that are citing the original article
• Try an author search from a good article
10 April 2008
Copyright: Ganesha Associates
2008
54
Practical activity 3a - What are
main the differences between…
•
•
Total duration - ca. 1 hr.
Select three information resources that are relevant to your work from the list below:
–
–
–
–
–
–
•
•
•
Google
Google Scholar
PubMed
Scopus
Web of Science
...or feel free to choose your own list
Search each of them with a short keyword string appropriate for finding new articles
relevant to your chosen field of research
Identify the main reasons why the results are different in each case. Prepare a brief
Powerpoint presentation describing the strengths and weaknesses of the three
resources, to be given at the start of the next session.
The strengths and weaknesses of the different commercial databases is hotly
debated by librarians - see, for example, these two reviews comparing and
contrasting Google Scholar, Scopus and Web of Science and Web of Science and
Scopus. Do you think the issues raised by the authors are the most important issues
?
10 April 2008
Copyright: Ganesha Associates
2008
55
Practical activity 3b - What Web resources
would you build to support your research
• Total duration - ca. 1 hr.
• Read the article by Charles Godfray. Think
of some Web resources that would be
helpful for your area of research. Discuss.
10 April 2008
Copyright: Ganesha Associates
2008
56