Version 13 of the Human Protein Atlas, Clickable, Downloadable, and Nearly Complete, is Now Online

By John Otrompke

December 18, 2014 | A new version of a research tool described as the world’s first spatial index to the human proteome is expected to change the nature of drug development, and is already the source of some two external peer-reviewed research papers on average every day, said Professor Mathias Uhlén, at a press conference on Nov. 6 announcing the release of version 13 of the Atlas.

“I would be very surprised if all pharmaceutical drugs 30 years from now have not been designed using the Protein Atlas in some way,” said Uhlén, who is heading the project.

Version 13 of the Human Protein Atlas was released, along with a poster, by the Swedish Royal Institute of Technology at a press conference November 6. It is the first tool to show users where the proteins are in different human tissues, said Uhlén, professor of biotechnology at the Royal Institute of Technology in Stockholm, and director of the program which has produced the Atlas. The Atlas is free to use.

The Atlas shows not only where in the body the proteins are, but can show, at the single-cell level, where in a cell a protein lies. The tool indexes some 13 million high resolution digital images, all of them annotated by a team of pathologists at Lab Surgpath, Mumbai, India.

The developers created a set of 20,000 antibodies used to test for the presence of each protein in a given cell. The antibodies are used to stain slides containing 3,151 human tissue samples—1,340 samples of normal tissue, and 2,123 samples of cancerous tissue— said Cecilia Lindskog, site director of the Uppsala site of the Atlas project, and a researcher in the Department of Immunology, Genetics and Pathology at Uppsala University. The results of the staining pattern are also compared with data from previously published literature.

“Today, all the pills in the world are directed at 620 proteins, which is 3% of all human proteins. We believe that by characterizing the 97% of proteins toward which no drug has been directed, scientists will find new targets for drugs, and also diagnostics,” explained Uhlen. “For example, if you are trying to make a drug directed toward a protein in the kidney, it could be very interesting to know if that protein is also in the liver or brain, and answer the question, ‘Does that give you a side effect?’” he added.

The Atlas integrates both genomic and proteomic data on 83 cell types, corresponding to 44 organs. The proteins are analyzed using either immunohistochemistry or confocal microscopy. Version 13 of the Atlas, which was released in November, includes 27 chapters on all the major organs. The data is also complemented with mRNA data from 32 normal organs, from the genomic browser at the University of California at Santa Cruz (Ensembl version 75.37).

“Most of the data in version 13 of the Atlas can be downloaded,” said Lindskog, a researcher at Uppsala University, who is helping to develop the Atlas. (Lindskog also contributed data on protein expression in pancreatic cells from diabetic patients). “That includes protein expression data, antigen sequences, Western blot data for antibodies, protein array data for antibodies, RNA-Seq data, and external references such as UniProt identifiers, and more,” she added.

The Atlas was also the subject of a Keystone symposium in April of this year, as well as an article in Nature on May 29, entitled, "An Atlas of Expression." The website has seen visitors from 150 countries, about 35% of whom are from the U.S.

Work was begun on the project in 2003 after the human genome project was completed, and the protein atlas went on-line in 2005. However, the developers still intend to have an annual release for the next ten years, as there are still some 3% of the human tissue types that are not represented, such as the retina or the taste buds, for example, Lindskog said.