Embedding machine-readable proteins interactions data in scientific articles for easy access and retrieval
Correspondence: (Login to view email address)
- 'L Galvani' Center for Biocomplexity, Universita' di Bologna, Via San Giacomo 12, 40126 Bologna, Italy
- Biocomputing group, Department of Experimental and Evolutionistic Biology, via San Giacomo 9/2, 40126 Bologna, Italy
PDF (275.2 KB)
- Document Type:
- Manuscript
- Date:
- Received 29 September 2008 14:22 UTC; Posted 01 October 2008
- Subjects:
- Bioinformatics
- Abstract:
Extraction of protein-protein interactions data from scientific literature remains a hard, time- and resource-consuming task. This task would be greatly simplified by embedding in the source, i.e. research articles, a standardized, synthetic, machine-readable codification for protein-protein interactions data description, to make the identification and the retrieval of such very valuable information easier, faster, and more reliable than now.
We shortly discuss how this information can be easily encoded and embedded in research papers with the collaboration of authors and scientific publishers, and propose an online demonstrative tool that shows how to help and allow authors for the easy and fast conversion of such valuable biological data into an embeddable, accessible, computer-readable codification.
Discussion
- Votes:
-
5 votes
- Comments:
-
2 comments
Dear Dr. Chenette,
thank you for your comment, and for your good question about such a hot topic. Diversity and dynamics of protein-protein interactions (PPIs) is an important and non-trivial issue, and we agree that discrimination of PPI types deserves more attention, also in the perspective of the machine readability of such data. We think that the differentiation among several interaction types is already, and will become more and more, necessary and of great significance in the near future. One of the current problems, especially with high-throughput experiments is the fact that there is not an obvious way to discriminate between transient and sustained interactions.
As for what specifically concerns machine readability, we could propose and easily adopt an annotation system that includes a further level of detail, f.i.:
1. protein A interacting with protein B
1.1 general interaction (default)
1.2 obligate interaction
1.3 transient interaction
In this way, it is possible to enter this kind of experimental data, when present. - (Login to share with a colleague)
Additional information
- License:
- This document is licensed to the public under the Creative Commons Attribution 3.0 License
- How to cite this document:
-
Tieri, Paolo, Termanini, Alberto, Fariselli, Piero, and Franceschi, Claudio. Embedding machine-readable proteins interactions data in scientific articles for easy access and retrieval. Available from Nature Precedings <http://hdl.handle.net/10101/npre.2008.2337.1> (2008)
- Version info:
-
Other versions of this document in Nature Precedings
None.
Other versions of this document elsewhere on the web
None known.
Emily Chenette on 08 October 2008 20:04 UTC
Dear Dr. Franceschi and colleagues,
Thanks for sharing a very interesting and timely document. I agree that protein-protein interactions in journal papers must be machine-readable for the data to be useful in the long term. I am curious, however, how you’d ask authors to discriminate between transient protein-protein interactions (for example, a kinase that temporarily interacts with a substrate) and sustained protein-protein interactions (e.g. scaffolds, or transcription factors with co-activators). It seems to me that these could be considered two discrete types of interactions, with different functional consequences. Is there an obvious way to discriminate between the two? Do you think that such differentiation is necessary?