Further Information

1
British Society for Proteome Research/European
Bioinformatics Institute Meeting From Proteins
to Systems 27-29 July 2005 Hinxton Hall
Conference Centre, Wellcome Trust Genome Campus,
Cambridge Proteomics Networks on the World Wide
Web Ruth McNally Peter Glasner Centre for
Economic Social Aspects of Genomics (CESAGen),
Cardiff
Table 1 shows that the network contains a variety
of different types of websites, both in terms of
how they are classified on proteomics society
website links lists, in terms of their top
level domain. It is notable that there are no
.gov nodes. Previous research suggests that the
absence of .gov websites which are typically
sources of information about regulation funding
is typical of a network in its early stages.
vi It will be interesting to monitor this,
especially after the European Proteomics
Association (EuPA) becomes active following its
first General Council Meeting at HUPOs 4th
Annual World Congress in Munich in August (which
appears in the network as hupo2005.com).
CESAGen The Centre for Economic Social Aspects
of Genomics (CESAGen) is a specialist social
science research centre based at Cardiff
Lancaster Universities in the UK. Founded in
2002, it is funded by the Economic Social
Research Council (ESRC), is part of the ESRCs
Genomic Network. i
One of CESAGens Flagship Projects studies
proteomics using methods concepts from the
field of science technology studies (STS). ii
This poster is based on our experiments to map
monitor proteomics on the WWW using the
IssueCrawler, a software tool developed by
govcom.org. iii
Table 1 Nodes in the Network
The IssueCrawler The researcher supplies the
IssueCrawler with the URLs (website addresses) of
at least two Starting Points. The IssueCrawler

Classification of node (based on classification in proteome society website links lists) No. of Nodes Top level domain of the website address Top level domain of the website address Top level domain of the website address Top level domain of the website address
Classification of node (based on classification in proteome society website links lists) No. of Nodes .org .com .edu .ac .net, . info or country code
Proteomics Society 21 10 11
Mass Spectrometry Society 19 2 3 14
Commercial Supplier/Product 8 8
Congress/Meeting/Conference 11 2 4 5
Journal/Book 12 4 6 2
Resource/Tool/Course 7 1 3 3
Electrophoresis/Separation Society 5 2 1 2
Research Centre/Lab. 3 1 2
Other Society 3 2 1
Research Initiative 2 1 1
Totals 91 24 21 8 38

1. Crawls the Starting Points retrieves the
   outward links.
2. Performs co-link analysis on the outward links.
   Co-links become the network nodes.
3. Analyses the direction and intensity of linking
   between the nodes.

IssueCrawler results can be used to

1. Visualise a websites politics of association
   The decision to hyperlink to another website is a
   conscious decision the enactment of hyperlink
   diplomacy. iv Which websites are active
   linkers? What kinds of websites do they ( dont
   they) link to?
2. Simulate websurfing Issue Crawler maps simulate
   journeys that could be taken by someone surfing
   the WWW from the Starting Points. For the surfer,
   hyperlinks may be construed as a (dynamic)
   recommendation that accords the linkee status.
   Co-linking indicates shared recognition.
3. Locate networks Which websites are in the
   network? Which are out? How does it change over
   time?

Of the proteomics societies whose websites were
used as Starting Points, only the website of
Korean HUPO failed to pass the co-linking
threshold for inclusion in the network. The
research initiative nodes are the websites of
HUPOs Brain Proteome Project (hbpp.org) HUPOs
Proteomics Standards Initiative
(psidev.sourceforge.net), which is the only .net
website in the network. The websites of the 8
commercial supplier/product websites are typical
of .com websites of commercial organisations in
that they do not link to any of the other network
nodes. At the other extreme, .org websites
typically display a more promiscuous linking
style, in this network exemplified by the
following proteomic society websites
swissproteomicsociety.org, bspr.org, dgpf.org
hupo.org which, respectively, link to 65, 51, 32
23 other nodes. With regard to visible politics
of association, the hupo.org website only links
to proteomics societies which are affiliated to
HUPO, Italy HUPO (hupo.it) the Italian
Proteome Society (ipsoc.it) do not link to each
other.
Method Starting Points URLs of the websites
of 22 proteomics societies Crawl parameters
Crawl depth 1 Iterations 1 Co-link analysis
by page Starting Points not
privileged Date 18 July 2005
Figure 2 Cluster map top 20 nodes node size is
relative centrality (inlinks outlinks)
red nodes give receive links green nodes only
receive links
Figure 1 Cluster map all nodes node size is
relative to inlinks from the network node
colour corresponds to top level domain of website
URL (see legend panel)
Figure 2 maps the 20 most significant nodes in
the network. In contrast to Fig 1, node size in
Fig 2 represents outlinks to other nodes as well
as inlinks received. Red nodes give receive
links from the other nodes in Fig 2, whereas
green nodes only receive links. Note the much
reduced size of asms.org the other green nodes
compared to the red nodes. The size location of
swissproteomicsociety.org reflect its centrality
in the network, which is the result of its
sociable linking style.
Results Discussion The Issue Crawler identified
a heterogeneous network containing 91 nodes that
collectively host 100 co-linked webpages. There
are 431 directed linkages within the network,
but, as Figures 1 2 illustrate, these are not
evenly distributed.
For more information about this research see our
forthcoming Viewpoint Article in Proteomics,
Issue 12, August, 2005. Acknowledgements This
research is supported by the ESRC at CESAGen,
Cardiff.
Figure 1 is a cluster map where the size of a
node is relative to the number of other nodes
that link to it. v hupo.org, the website of
the Human Proteome Organisation (HUPO), is the
2nd largest node in the network. However, the
largest node is the website of the American
Society for Mass Spectrometry (asms.org) (15
inlinks), which implies that this website enjoys
the highest level of status/recognition within
the network. However asms.org does not repay the
compliment it links to only one node imss.nl
the website of the Intl Mass Spec Society,
implying that its politics of association on the
WWW are rather different to those of the
proteomics society websites that collectively
assembled the network nodes. There is a cluster
of mass spectrometry society website nodes on the
top right hand side of the network, near
spectroscopynow.com.
Notes i www.cesagen.lancs.ac.uk
www.genomicsforum.ac.uk/ESRC_Genomics_Network/inde
x.php ii www.cesagen.lancs.ac.uk/research/projec
ts/proteomics.htm iii www.govcom.org/about_us.ht
ml iv Rogers, R. and Marres, N. 2000, Public
Understanding of Science 9 1-23. v Cluster map
software is ReseauLu, by Andrei Mogoutov,
www.Aguidel.com vi Rogers, R. 2004,
Information Politics on the Web. MIT Press.
Further Information
Tel [phone] McNallyR_at_cardiff.ac.uk
www.cesagen.lancs.ac.uk