EMBL’s European Bioinformatics Institute

The beginnings of UniProt go back to 1986, when Amos Bairoch started developing programs for protein research to complete his master’s thesis at the University of Geneva. While distributing his sequence analysis software with the PIR Protein Sequence Database (PSD), which grew out of the world’s first collection of protein sequences chronicled in a series of atlases from 1965 to 1978 by Margaret Dayhoff, Amos ran into difficulties with cross-referencing, annotations and vital characterisation information due to PIR-PSD format.

Inspired by EMBL’s Nucleotide Sequence Data Library, Amos released ‘PIR+’ in a similar format, later renaming it Swiss-Prot. In 1986 he distributed the first release of the database freely using US BIONET – Swiss-Prot version one contained 3900 sequences. Today, the UniProt Knowledgebase offers over 63 million.

EMBL’s data library and Swiss-Prot were a natural fit, and led to a lasting collaboration. EMBL agreed to distribute the database and participate in Swiss-Prot’s maintenance, and in 1987 Dr Rolf Apweiler, now Director of EMBL-EBI, became the first person outside of Geneva to work on the project, and led the Swiss-Prot group in the UK between 1994 and 2014, when Alex Bateman took on leadership of the group at EMBL-EBI.

The challenge of funding open data

Until 1993 Swiss-Prot was funded by the Swiss Government, the University of Geneva and EMBL member states. But in the mid 1990s some of its funding streams came under threat. Meanwhile, the PIR-PSD had been funded by the National Institutes of Health (NIH) and came under the directorship of Cathy Wu at Georgetown University in the late 1990s. The EMBL-EBI Swiss-Prot group, in collaboration with the Geneva Swiss-Prot group and the Georgetown PIR group, applied to the NIH for funding, and this successful bid ensured the protein data resource could continue to be freely available to the global research community.  

As a result, Swiss-Prot became part of the Universal Protein Resource Consortium (UniProt) when it launched in 2002. The SIB, PIR and EMBL-EBI continue to this day to collaborate on the project’s expansion and distribution.

An international hub for talent

Swiss-Prot has continually expanded its curation and adopted new technologies to optimise annotation. “At first we typed everything manually on these first-generation computers, with flickering green lights resembling something from an old Star Trek movie,” recalls Claire O'Donovan, team leader for Protein Function Content at EMBL-EBI, who has curated Swiss-Prot since 1993. “That way of working would be completely unrecognisable to some of our newer team members.”

Collaborating with over 100 people demands regular engagement, with daily Skype calls, scheduled weekly videoconferences and yearly retreats at each site. For the UniProt Consortium, there is a lot of interaction between teams and occasionally people move between partner sites.

“Having such a diverse team leads to a lot of exchanges and better decisions in the long run,” says Maria Martin, team leader for Protein Function Development at EMBL-EBI. “Our teams cover all areas of expertise, so there is always someone to ask for help – and that is a big part of what makes UniProt work so well.”

“The physical distance between us doesn’t always make things easy, but being invested in what we do helps us rise above such limitations and work together to anticipate future needs,” says Sangya Pundir, UX Manager in the Protein Function Development team at EMBL-EBI.

“UniProt is one of those projects where even after you leave, you will always look back and want it to succeed,“ says Claire. “It is an amazing resource, and we are all lucky to be a part of it.”