The authors have declared that no competing interests exist.
This July,
What proteins can be found in a cell? How do protein complexes form, and why? How do gene families evolve, and what drives both gene duplications and epigenetic modifications? How is bioinformatics influencing personalized treatments of cancer cases? These four essential—and yet unanswered—questions were put forward to the ISMB 2011 audience by Professor Dr. Alfonso Valencia as the icebreaking launch pad of his keynote talk to challenge the community to develop a concerted effort in the field.
In the past few years, it has become evident that alternative splicing is one reason why human genomes can produce so much complexity with so few genes
How do proteins manage to distinguish the right binders (cognate interaction partners) from the wrong ones? To address this second question, Valencia reported a high-throughput docking experiment, showing that physical docking can often identify correct binders by predicting the interaction partners and the organization of the interaction surface using the distributions of the docking scores for over 1 billion of complex models generated
But, why do cancer cells accumulate structural variations? Is tumor progression analogous to species evolution? Is gene duplication a positively selected process, or is it an inevitable consequence of the mechanism of DNA replication? Both chromatin structure and DNA replication dynamics play a role in eukaryotic genomic evolution, and replication induces cellular stress, with exposed single-strand DNA leading to DNA damage. In the third part of the talk, Valencia put together DNA replication dynamics
Subsequently, the talk went deep into the title topic, highlighting a recent review in the field
Last, but not least, Valencia underlined the contribution of Spain to the International Cancer Genome Consortium (ICGC)
The final take-home message was a call to stimulate the interchange of methods (software) and data (validated sets) within the scientific community, promoting in-harnessed collaborations across research groups. As Valencia said, “there is no gain by developing these systems in isolation or implementing only everyone's own software”.
Did Valencia achieve his purpose of challenging the audience? According to Bryant, “I was definitely interested right from the start. One of the things I liked about the talk was that he presented information I had not known about, and it really got me thinking”. If we consider the high number of questions in the discussion following the presentation—I counted nine in 11 minutes—the high impact it had on the attendees becomes evident. A good talk has the audience making guesses and I felt that Dr. Valencia did that well.
To sum up, recent developments in molecular biology aided by computing are paving the way in the era of genomics medicine, and new opportunities are emerging to detect genetic events leading to further progression of cancer. In my opinion, it will change the assumptions under which conventional treatments such as radiotherapy or chemotherapy are applied today to each patient, where the precise nature of genetic damage and the mutations involved are not yet well known. Thus, facing the up-to-date challenges expounded by Valencia may be considered as the next stepping stone to the utilization of personal genomics in forthcoming individualized cancer treatments.
Many interesting lectures were given at the ISMB 2011 conference in Vienna. In my opinion, one of the outstanding sessions in the conference was the work dedicated to understanding the mysterious role and function of proteins encoded by chimeric transcripts, which was presented by Milana Frenkel-Morgenstern, a post-doctorate fellow in the CNIO in Madrid, Spain. Alternative splicing is thought to influence more than 70% of human genes and has a major contribution to both transcriptomic and protemic diversity. It has been shown to have a role in several genetic diseases as well as in cancer development. Chimeric transcripts may be generated by
By using the above genome-wide detection of chimeras and their functionality analysis, many specific events of special interest could be identified. For example, the chimera resulting from the fusion of the transcription repressor (Ctbp1) and transcription factor-3 (TCF3) produces a dominant negative protein that deactivates transcription. Another example showed the incorporation of signal peptide and transmembrane domain resulting from the fusion of solute carrier family 22 member 6 protein (Slc22a6) and thioredoxin domain-containing protein 12 (Txndc12).
I think that this talk raised an important discussion about the consequences of generation of chimeric proteins in cells. These chimera are likely to have substantially different functions than the original native proteins. This work indicates that it is feasible that these chimeras could have acquired specific functions and that they might exert dominant negative effects due to the absence of certain functional domains and therefore might compete with functional wild-type proteins.
Generally, I found that this talk was a good illustration of how experimental biology can benefit from computational approaches. The ISMB conference encourages the usage of advanced computational methods that resolve biological problems, which I believe was also exemplified by this talk. My personal feeling is that the work presented by Dr. Milana Frenkel-Morgenstern illustrates how important and valuable the use of computational methods is along with high-throughput screening for the analysis of protein functionality and characterization, and how they could contribute new hypotheses and insights for answering biological questions.
Professor Brunak (Technical University of Denmark and University of Copenhagen) presented the first talk in the BioLINK special session at ISMB 2011 on how to utilize a systems biology approach to look at diseases phenotypes. The BioLINK session was organized by Christian Blaschke (Bioalma, Spain), Lynette Hirschman, (MITRE, United States), Hagit Shatkay (University of Delaware, United States), and Alfonso Valencia (Spanish National Cancer Research Centre, Spain).
With the BioLINK session focusing on data integration and interoperability across the computational, biological, and medical fields, Dr. Søren Brunak reported new gene–disease associations that have been discovered by integrating phenotype data with molecular data. In his talk, Dr. Brunak demonstrated how his group utilized electronic health records (EHRs) of Danish patients to extract patient-level phenotypic data. Unlike the United States' recent Medicare and Medicaid incentives, the Danish government launched their national strategy for EHRs much earlier, in the 1990s. Fortunately, there are still a few health care providers in the US, such as Marshfield Clinic, that have multiple decades of clinical data in the form of EHRs. These EHR datasets across the continents make it possible for future cross comparison and validation of the findings by Dr. Brunak's group.
A limitation indicated in Dr. Brunak's talk is that the connections between the molecular entities (for example, genes) to diseases are only at an aggregated level. In specific, the molecular data were text-mined from the OMIM database and other scientific literature. As such, patient-level variations, which are the crux of personalized medicine, were lost. As Dr. Brunak pointed out, molecular measurements from a biobank of patients can potentially solve this problem. The well-curated biobank with links to EHRs can be used to characterize the genotype-phenotype variation at the patient level. And several well-established biobanks in the US, such as BioVU at Vanderbilt University and the Personalized Medicine Research Project (PMRP) at Marshfield Clinic, can offer help.
EHRs remain a rather unexplored, but potentially rich, data source for most computational biologists. Dr. Brunak's avant-garde work represented the forefront of translational bioinformatics, which is defined as “the storage, retrieval, analysis, and dissemination of molecular and genomic information in a clinical setting”. Both the International Society of Computational Biology (ISCB) and American Medical Informatics Association (AMIA) are actively promoting translational bioinformatics.
The disciplines of bioinformatics and medical informatics are closely related and they can be synergized to achieve the goal of personalized medicine (
In summary, the convergence of bioinformatics and medical informatics can open new paths of exploration for personalized medicine. Dr. Brunak's talk was a good indication that more future joint activities by both ISBM and AMIA will benefit both current and future generations of biomedical informatics professionals.