Want to keep learning?

This content is taken from the The University of Glasgow's online course, Cancer in the 21st Century: the Genomic Revolution. Join the course to learn more.

Skip to 0 minutes and 14 secondsHello. We had a quick look in the first video at Google, but you could see we were still finding lots of information that was coming from pressure groups, charities, other bodies, and just general commentary all over the internet. What we can do to actually facilitate getting to quality information is to use a database. One of the best databases to demonstrate to you because it's available free to anyone on the planet is the one called PubMed. Just googling the word "PubMed" pulls up the home page. And you can see that this actually comes from the National Institutes of Health within the US National Library of Medicine. It's made freely available for anyone on the planet to use.

Skip to 0 minutes and 57 secondsIt only indexes quality information from peer-reviewed sources, usually journal articles but not always. The main system that you can search is PubMed. And we could just throw a word in here as we did with our original search on Google. You can see it's suggesting whether we would like to look at a more specific type of cancer, but let's just run the search as it is with a very general term. Now you can see we found loads and loads and loads. If we come down and have a look, you can see what PubMed's actually done with the word that you typed in.

Skip to 1 minute and 27 secondsWhat it's done is it's said, I'm looking for the word "cancer," and I'll tell you about any articles that have the word "cancer" anywhere in the record in all the fields, whether it's in the author, the title, or the abstract. It sensed that I'm looking for the word "neoplasms" within all the fields. And you might think, well, where has it got this term "neoplasms" from? And that's because PubMed is quite clever, and in the background it has mapped the term that you've put in to the MESH term that it has. So I hear you say, what is a MESH term?

Skip to 1 minute and 58 secondsIt actually stands for Medical Subject Heading, and these are added value that you get by using the PubMed database, because what they do is they help control for all those things like author variation and language, differences in spelling, perhaps there's been a typographic error in the original document, or there's a geographical variation in how your topic is talked about. How the MESH terms get assigned to records is that the indexer will add them to the record about the article-- for example, this gene expression correlation for cancer diagnosis one. Someone will have looked at that and said, it's predominantly about cancer.

Skip to 2 minutes and 34 secondsThe accepted MESH term for anything relating to cancer is "neoplasms," so I'm going to add the term "neoplasms" to that record within the MESH term field. Now for you as a researcher, that's great, because you can come along, search on the MESH term, and know that you're getting everything that's on that topic and that covers that topic in quite an in-depth way within the article regardless of how the authors talked about it. If you want to see the difference, though, what we can do is just search the MESH headings. So let's run our cancer search within that MESH heading field. We're not looking at articles here. We're just looking at the words that are used.

Skip to 3 minutes and 14 secondsIt's quite helpful to do this when you first start. So even if you want to go and do a Google search, or a search within another system doesn't have these MESH terms, you can run a search in here to see what vocabulary other people on the planet might use, or how you might want to describe your topic. So let's click on "neoplasms." And actually, yes, that is what I mean by the term "cancer." One of the other clever things we can do is look at these subheadings.

Skip to 3 minutes and 38 secondsAnd the indexer, when they put the information about articles on, will have gone, well, not only is this article about neoplasms, it's actually about the genetics of neoplasms, so I can click a subheading. Now at this point, I've clicked a box, but nothing has happened. There's nothing in my search box, and there's nothing in my search builders. So how do I actually get the system to look for information for me? I have to click on 'Add to Search Builder'. And you can see, it puts this in in quite a complex way.

Skip to 4 minutes and 4 secondsIt's far easier just to look the word up, tick a box and say "add" rather than try and remember, oh, it's a quote and it's a slash and it's a box and a MESH and that sort of thing. You'll then say search PubMed for information. And you can see we've dropped from 2 million very generic articles right down to 279,004. If we then want to look for a particular article-- because some of these don't look particularly relevant for genomics research-- let's delete this, and run another search, and say, well actually, I'm quite interested in the role of LIMK2, and let's search. And it tells me that I've got 102 records.

Skip to 4 minutes and 45 secondsBut if you remember, what we found previously is that different people type these things in different ways. So let's just try a few of those ways, and see if it makes a difference to what we find.

Skip to 5 minutes and 3 secondsLet's run the search. And you can see we go up to 107. So we've found five extra articles, and one of those could be actually key. Looking down at the first few, I can also see that someone has typed it as LIM kinase-2 as a full phrase. So let's now add that in, as well, and see if that makes a difference. And you can see we've gone up to 112. Now at this point, you could argue, well, LIM kinase-2-- how is that relating back to the cancer search I did before? Well, at the moment, it doesn't. But what you can do is click on 'Advanced', and you can see all the searches we've done before.

Skip to 5 minutes and 40 secondsSo what we can now do is say, well, please take those 112 records, add them into the search builder-- which is this bit up here. And what I want to do is combine that with my genetics search. And again, make sure they're linked with "AND." So PubMed's now only going to give me information that has one of these terms in it, and has the neoplasms and genetics MESH heading. Let's run a search. And I'm down to a nice seven results. There are various limits here on the left which I could use. So I could come in and have a look at more. I can have a look at the ages of my participants within studies.

Skip to 6 minutes and 22 secondsI can see what article types there were. Do I want clinical trials? Do I want review articles? So within PubMed, one of the advantages as well is, because it's a medical database, you have clinical limits that you can apply very easily just to filter your information. You can then click on the title to get to more information. And this one's actually made available open access. And increasingly, grant bodies are saying to their researchers that they must make the results of their research available to people on the planet, which is the open access model. It's a condition of being given the grant. So within PubMed, you will start to see more and more of this open access information become available.

Skip to 7 minutes and 2 secondsYou can limit to open access full text within PubMed as one of those facets. Coming back to the PubMed home page, you can see that there are lots of helpful guides on using PubMed. There's quick start guide, which is very good just to print off and have as a handy reference. There's lots of tutorials and videos. And I would recommend that you have a look through these. It's a very powerful tool, and you can use it at a very baseline level, as we have today. But you can also use it to save searches, give yourself a current awareness alert about new developments within the field. So I hope that's helped, and happy searching.

Using a scientific literature database (7.44)

Heather Worlledge-Andrews gives an overview of how to use PubMed, a very useful and freely available literature database in your research. You might find this useful in the following steps.

Share this video:

This video is from the free online course:

Cancer in the 21st Century: the Genomic Revolution

The University of Glasgow