[MUSIC].
Hello and welcome.
We'll be looking at searching PubMed
effectively.
PubMed is a government website, based at
the Natural Library of Medicine.
It contains over 23 million biomedical and
life science publications.
Searching it effectively, can be a very
useful strategy.
And it's also relevant, because the
mechanism for searching Pubmed, is
remarkably similar to the method for
searching nucleotide and protein sequence
databases.
[MUSIC]
Our objectives include, searching by
author.
For instance, finding papers written by a
particular individual.
Either by using the last name format, or
by using one or two initials.
We will also look at how to search with
field tags.
In other words, limiting your search, to
very specific fields in PubMed.
We will search by subject, and combine
searches with subject, and an author name.
That is a Boolean operation.
[MUSIC].
We will look at the MeSH database.
MeSH, has a controlled vocabulary, that
allows for more effective searches of
PubMed.
We will use filters.
Filters, help narrow down to certain types
of papers.
We will also examine the truncated
searches, which
allow for searching for variations of a
particular work.
[MUSIC]
First, let's go to PubMed.
There is an address for PubMed, at the
National Library of Medicine.
However, pubmed.gov, that is P-U-B-M-E-D
.gov is a working alias.
So, pubmed.gov will take you to the NIH
website where PubMed is hosted.
Let's use the word, greider, and search.
You will see some papers that include
an author named Greider, which I just
highlighted.
You will also see some papers, where
Greider, is not listed as an author.
Carol Greider is a nobel prize winning
researcher of telomeres.
So perhaps Greider was mentioned in
[SOUND] that particular paper.
We'll look at a way to avoid those kinds
of results shortly.
The last name Greider, finds 223 results
in PubMed.
If I want to limit further to specific
authors with the last name Greider.
I can add a first initial.
So, Greider C, would narrow down to 102
papers.
You can see that Carol Greider's second
initial is W.
A one initial search, will find all
variations
of C, so C followed by any second initial.
Or just c by itself.
[BLANK_AUDIO].
I could further narrow, by adding the
second initial.
You see, that narrows the search down to
94 papers.
So basically, last name is the broadest
search.
One initial is a little more specific, and
two initials are the most specific.
Some authors use more than two initials.
However, PubMed limits searches to two
initials, and
PubMed actually only enters two initials
for any author.
Okay, now I mentioned, there is a way
to avoid matching words outside the author
field.
I'm going to go back to searching for
Greider, last name only.
If you noticed, the first paper did not
have Greider as an author.
So I'm going to open that paper now.
When you open a record it's going to show
the abstract.
So here's the abstract.
And you can see that it says that the 2009
Nobel prize
in medicine was awarded to, there it is,
Blackburn, Stostak, and Greider.
So her name is in the abstract, and not in
the author field.
The author field lists only one author,
and that is Wang.
So, if I want to limit the search to only
the author field.
The way to do that, would be to type,
greider(author).
[BLANK_AUDIO].
Now, I only see those papers, where
greider is in the author field.
I eliminated that result, and others,
where greider was in the abstract.
Basically, by putting brackets around the
field name, I
can limit a specific term, to a particular
field.
There are acceptable abbreviations.
For instance, if I had typed, bracket au.
Au, is an accepted abbreviation for
author, I would get the same result.
There are a list of fields available
that are searchable, including the author
field.
Now, suppose I wanted to search by
subject.
There are several ways to do it.
And usually one of the easiest ways to do
it, is to simply type in the subject.
So I am going to type in, RNAi.
RNA interference, is a very, very popular
search topic.
And, a very important topic in life
science.
RNAi, was famously studied by Craig Mello
and Andrew Fire.
So, let's say that I want to get papers on
RNAi, by Craig Mello.
I can type in,
RNAi And,
mello(author).
Now, what you might notice, is that and is
capitalized
and for all searches, you should
capitalize the Boolean operators.
Boolean operators are, and, not, and or.
This gets me papers that are about RNAi,
and are
written by an author named Mello, which
gives me 27 papers.
Some of them have, RNAi in the title.
Some of them don't.
So, if I want to further limit RNAi to the
title field.
After RNAi, I would type (title), and
you'd see the list
is now narrowed to ten papers that have
RNAi in the title.
You can see RNAi bolded in the title field
of each paper.
Okay.
So we can see, how to limit to very
specific fields.
We can also see how to use the and
operator.
Be sure to use all capital letters for all
Boolean operators.
So that it doesn't look for the word, and
as a search term.
There is another way to search PubMed by
subject.
That involves, using the MeSH database.
MeSH which stands for medical subheadings,
uses a controlled vocabulary.
It allows for more consistent complete
searches.
Some writers, use different words to
describe the same phenomenon.
The controlled vocabulary is used to index
each specific paper consistently.
There is a very good video on
YouTube, provided by the National Library
of Medicine.
And it is available at this address, and
you can view that
video called, using MeSH to build a better
PubMed query, for more information.
[BLANK_AUDIO].
Quickly though, if I want to search RNAi,
instead
of PubMed, I would now search the MeSH
database.
And you see, that the preferred term,
again, this
is a controlled vocabulary, is RNA, column
small interfering.
If I want to search PubMed, I would simply
add it to the search builder.
And it may be even add.
And, Mello (author) and then click search
PubMed, I actually
got 25 papers, some different from the
ones I got using the same
search, with just RNAi, that's because
sometimes the term
RNAi, might not have been used in those
particular papers.
But those papers are about RNA
interference.
Therefore, I get those additional papers,
using the control vocabulary.
You might also notice that on the left
menu, are a series of filters.
Those filters include article type, so if
I want only review
papers, I would click Review, and that
would narrow it down further.
A popular one is text availability.
Most papers, when you open the record, it
only shows the abstract.
But most papers have a link to full text.
And sometimes the full text, comes at a
cost from the publisher.
Sometimes it's free, and very often, you
would want free full text, why not?
If I click the, free full text available
filter, then I narrowed it down to 16
papers.
Now, when I click that first paper.
I know that there is a link that reads,
cell
press, and it says, free NPMC in the upper
right corner.
When I click that link, I am going to be
able to download the
entire paper for no cost, because I chose
that filter of free full text.
That's very useful, if you are not at an
institution that subscribes to
a bunch of specific journals that will
allow you to have access to them.
The truncated search could be very useful
if there are variations of a word, for
instance, if you're unsure of spelling, or
if there is a singular plural issue.
An asterisk can represent the rest of the
word.
The asterisk can only come at the end of
the word, and not at the beginning or the
middle.
Immunoglo asterisk, will hit both the
singular immunoglobulin, and the plural
immunoglobulins.
You don't need to search twice.
[BLANK_AUDIO].
It's also very useful for an author name,
if you don't know the spelling.
However, there is a limitation.
It only shows the first 600 alphabetic
variants.
For instance, if you are using the name
McDonald, and,
you're not quite sure if it's McDonald or
McDonald's son.
Then it is better to type, M-C-D-O-N star
than just,
M-C-D star, put in as much of the spelling
as possible.
An example of that particular type of
search
is to type, immunoglo star and smith (au).
I now get papers on immunoglobulin, and
immunoglobulins.
The last topic to cover, involved the date
fields.
Both EDAT for entry date, meaning the date
that that paper was entered into PubMed.
Or PDAT, publication date, meaning the
date on the journal itself.
For instance, the January 9th issue of
Science.
Ways to limit to specific dates, include
year.
2014, will limit you to the entire year
2014.
Or 1998 for the entire year 1998.
Year/month, so, 2014/03,
will give you, all of March 2014.
Finally, the specific date.
2014/03/17, gives you March 17, 2014.
It's most commonly used in a range,
in a range of dates, 1998/02:
the colon indicates a range.
2013/09/27.
That translates to February 1st.
Since it's at the beginning of the range,
you get the whole month of February.
February 1st, 1998 through September 27th,
2013.
An alternative type of dates search is,
"last 19 days"(edat).
Of course, the number can vary.
So, for example, if I had heard about a
paper that was submitted to nature
within the last ten days, I would type,
"last ten days"(edat).
I know that it was submitted within the
last ten days.
Not with a publication date within the
last ten days.
And, nature(journal).
That limits the search, to the journal
nature.
[BLANK_AUDIO]
Well, here's a good lesson.
I still have a filter activated.
The filter, is free full text available,
and nature
doesn't have free full text, so there are
no results.
I have to clear that filter, because those
filters persist.
So, if you use those filters, be very,
very careful.
They stay on in subsequent searches unless
you turn them off.
Now, if I want to limit to a particular
date range, let's say we know that
it was toward the end of 2011 or
was the beginning of 2012, somewhere
around that time-frame.
So let's go for all of November 11.
Two, well let's get all of February 2012.
So, let's type, 2011/
11: 2012/02[edat].
Let's say it's a
science paper.
So, capital AND science(journal).
And, it was something about DNA, so I'll
put in DNA.
I found, 37 papers, that were in the
journal Science, that were
about DNA, and sort of toward the end of
2011 or beginning of 2012.
Maybe I could now sort through, and find
the paper I was thinking about.
[MUSIC].
This concludes our exercise on PubMed, and
I think the best
way to really learn it, is to go to PubMed
and
play around a little bit with some of
these field tags
and Boolean operators, and see what you
come up with, good luck.
[MUSIC]
[BLANK_AUDIO]