The Automated Function Prediction 2006 meeting will have a discussion panel, with our keynote speakers answering audience questions.
We have decided to solicit questions from participants prior to the meeting, out of which a few will be picked for the panel session.
Any questions relevant to the computational prediction of the function of biological molecules are welcome.
OK, I'll go first:
- What are the main challenges to creating a Critical Assessment of Function Annotation (CAFA) contest, in the spirit of CASP and CAPRI? (Iddo Friedberg)
- How would metagenomics impact the field? (Iddo)
- How is structural genomics already impacting the field? (Iddo)
- description of function: are ontologies the only way to go about it? (Iddo)
- How can we have an expected level of function assignment for a genome depending on literature coverage? Example: PMID 15864286  (Christos)
- Derive a more sophisticated scale of function assignment/annotation consistency? What is a 'good' function prediction? Example: PMID 11864365  (Christos)
- Function prediction for non-coding or non-translated elements. Example of the first would be genomic regulatory elements, of the second would be RNA. (Iddo, Adam)
- What is the proportion of proteins that have multiple functions at the molecular level? no function at all? (troll warning) - the point being: what evolutionary path does a protein follow in terms of function? (a) Gain of function first and eventually duplication (2 functions at the same time) or (b) duplication, loss of function, gain of a new function (Martin)
- Do we need to include structural information into prediction models or does the primary structure already contain all the information we need to know to infer function? (Dani)
- Janssen P, Goldovsky L, Kunin V, Darzentas N, and Ouzounis CA. . pmid:15864286.
- Ouzounis CA and Karp PD. . pmid:11864365.