|
Can you imagine trying to do your research without GenBank and PubMed?
These public databases of sequences and literature citations have become
such essential tools for the research scientist that it's hard to imagine
life without them. Open Access Now talks to David Lipman, the man behind
both of these as well as his latest project – PubMed Central, a digital
archive of the full-text biomedical literature that will be a key component
of life in the Open Access future.
Integrating literature and databases
PubMed and GenBank are services run by the National Center for Biotechnology
Information (NCBI), part of the National Library of Medicine (NLM),
based at the US National Institutes of Health campus in Bethesda, Maryland.
NCBI’s Director David Lipman was there at the beginning, when NCBI
began pioneering database integration. “PubMed came out of two
streams that merged together,” recalls Lipman. “One was that
the NLM had been producing MEDLINE for many years as an electronic index
and bibliographic search service for the biomedical and life sciences
literature. MEDLINE contained the citation and abstract, with some keywords
and indexing terms. It was being made available online, usually by telephone-based
access and then via internet as well, for a small fee to recover certain
costs. About ten years ago NCBI started producing GenBank and providing
some CD-ROM-based services for DNA and protein sequences.”
“We started to integrate MEDLINE with the sequences, so that if someone
published a paper with a DNA sequence in it then we would have the corresponding
MEDLINE record integrated with the sequence record,” says Lipman. “That
was very popular and we kept increasing the MEDLINE coverage because
the people using the service found it very useful. The set started getting
bigger and bigger and then we thought about just including all of MEDLINE
and making it a free online service.” That’s how PubMed was born.
The NCBI made it available during a test period and then the NLM Director,
Dr. Donald Lindberg, decided to make it an ongoing free service. “We
showed that some of the costs could be cut by using the internet and
that by making it freely available it would be used a lot more,” Lipman
says. “Because PubMed was a free service offered by a dependable organization
– the NLM has been around since the Civil War – academic and commercial
groups felt confident incorporating PubMed into their own information
systems.”
PubMed demonstrated that the internet could make
information openly available more cheaply and that it would be more
widely used
PubMed was an instant success and within a short period of time its
usage had shot up 100-fold. “PubMed was very popular and widely used,”
says Lipman. “Two thirds of the people who access the system are not
academics – they may be healthcare professionals, teachers or students.
PubMed also creates a link-point for a lot of other kinds of information.
Because it’s there, and it’s free and it’s dependable, a lot of other
information providers use it in different and interesting ways - ways
that we never would have imagined when we first started.”
PubMed Central: a central archive
A few years ago discussions about Open Access to the literature led
to the creation of PubMed Central. “From the beginning, PubMed Central
really had two roles,” says Lipman. “One was to try to increase access
to the information and the other was electronic archiving. As electronic
journals became more important and digital versions of articles, with
more information, started to replace print as the definitive versions,
the issue of archiving came up. “For paper journals, archiving was always
done by libraries, so it was natural for the NLM to work with publishers
to provide digital archiving through PubMed Central. So we started PubMed
Central firstly to see if we could get the publishers to provide open
access through our system and secondly to provide digital archiving.”
“While Open Access remains controversial, the digital archiving side
of it is not,” Lipman says. The publisher provides the scientific content
tagged in an explicit format, and NCBI converts it to a common format
and makes it openly available. “If the technology changes, or there
are doubts about publishers’ stability, then we have a single format
for all this content so that we won’t have to spend outrageous amounts
of money to update everything,” explains Lipman. “The NLM has a long
history of active work in archiving. I think that one of the most important
things that we are doing is dealing with the issue of stable archiving
of digital scientific content. To that end, we have put a lot of work
into developing standard archiving formats, working together with academics,
publishers and university librarians.”
| |
“The current status of PubMed Central is that we have a number of largely
society publishers participating - some of them provide their content
immediately, and free, while others have up to 12 months delay,” says
Lipman. “We keep getting more participants, but it’s a slow process.
Usage has been increasing – it’s now several times what it used to be.
We have done nice work integrating PubMed Central into our overall Entrez
system, so that you can link more directly to our other factual databases
and, for example, find all the protein structures in a set of papers,
or find all papers associated with a taxonomic group like the metazoans.”
One of the most important things that we are doing is dealing with
the issue of stable archiving of digital scientific content
Another important aspect is how PubMed Central intersects with the
Open Access movement. “There are going to be new journals starting up
and one way to ensure the stability of this information is that there
be national archives like PubMed Central, so that even if the journal
is short-lived PubMed Central will continue to make it available for
many, many years to come,” explains Lipman. “We have already started
formal arrangements with groups in other countries to set up their own
archives to contain the same content.” NCBI is encouraging others to
have their own software systems, as there may be some advantages to
providing diversity and having additional automatic checking of content.
One of the critical aspects is that as the Open Access movement grows,
people are going to want to have national archives with long standing
commitments to preservation. And that’s where PubMed Central comes in.
Bringing papers to life Lipman predicts that the growth of PubMed Central
will bring new approaches to how data and literature are integrated.
“There are some obvious things that we can do now with the current PubMed
Central system that clearly add value. For example, we can integrate
a figure hidden away in a paper with factual databases of protein domains
linked to structures and so forth. So that after the paper is published
you could look back at the figure but see it actually updated with new
information about structure. Then you would really be able to look at
the paper and understand and extrapolate further what they were trying
to do.”
When we have Open Access, people will be able to create and use databases
and the literature in ways that will make both better
Lipman predicts an exciting future for further integration. “When people
really view the literature as being owned by the community and as the
tools get more sophisticated, then I think we will see people provide
links in their papers to the factual databases in such a way that the
paper is sort of alive. And setting up new kinds of databases, things
like image databases, will be almost trivial.”
“When we have Open Access, then more people can get at it and that’s
obviously better because more people can think about the science,” says
Lipman. “But when it’s a community property and people are creating
databases along with creating the literature, then the databases get
better. Sometimes it’s harder to interpret the databases independently,
but when you know that this paper is connected to a database you can
take advantage of that. And likewise, the papers can have a more dynamic
nature because they are linked to databases that are being continuously
updated.”
“We have built a lot of this into PubMed Central already. But I am
confident that as more people are working with it, it will start to
evolve in directions that we could not have predicted right now. People
will be able to create and use databases and the literature in ways
that will make both better. And creating better databases, for less
money, will allow better integration with the literature.”
Lipman sees the scientific paper as evolving over time as new discoveries
are made. “Open Access will speed up the evolution of the scientific
paper. When I see how much the biological databases have changed and
improved over the last two decades, it’s exciting to think about what’s
in store for the scientific literature in an Open Access future. I think
that the system will become much more powerful as more and more of the
literature becomes more open. I think the nature of the scientific paper
will change – and we don’t know what it’s going to turn into.”
www.pubmedcentral.gov/
|
