Implications of eScience for Science and Society: A View from Genomics

1. Implications of eScience for Science and Society: A View from Genomics Adam Siepel Assistant Professor Biological Statistics & Computational Biology Cornell University

2. Why is the Genomics Perspective Interesting? • Area of tremendous growth • Computing plays a central role • Of increasing interest to nonspecialists • Important in the commercial, academic, and government sectors • Complex concepts, large data sets, lots of noise • The typical molecular biologist is a good proxy for a “citizen scientist” vis-à-vis computer literacy

3. Current Landscape • Widely used public, web-based resources for researchers (NCBI Entrez, Genome Browsers, OMIM, PubMed, etc.) • Many academic analysis programs (of highly variable quality) • Some commercial analysis and visualization packages • Strong online/open-access publication movement • Decent educational resources for lay people: blogs, wikipedia, etc. (but unconnected) • Rapid growth in resources for personal genomics

4. Two Movements • Open-access publishing • Public Library of Science (PLoS): seven journals, rigorous review, high impact factors • BioMed Central (BMC): 190 journals (!), peer review, rapid turnaround, respectable impact • Both pay-to-publish, free distribution • Personal genomics • Most famous: 23andMe, Navigenics, deCODEMe • Consumer genotyping (500k-1M SNPs) from saliva samples for $1000 – $2500 • Electronic report of disease risk; in some cases geneology, nondisease phenotypes • Ethical, legal, and political controversy

5. Some Implications • More information for citizen scientists • Opening up of “black box” of scientific research; democratization of science • Opportunity to “write” as well as “read” • Increased dependency on electronic resources, eScience infrastructure

6. On the Horizon • Improved connectivity between resources at different levels: blogs, textbooks, wikipedia, browsers, research papers • Better access to underlying science from personal genomics gateways • Genotyping in a clinical setting (for diagnosis, guiding treatment); requires computational infrastructure! • Progress toward a more dynamic, interactive publication model • Steady progress in usability and integration of tools and databases

7. Also Coming (?) • Use of new genotyping technologies in forensics, paternity testing, etc. • A much stronger form of personalized medicine (every genome sequenced) • A completely different publishing model (?) • Analytical power tools in the hands of citizens (association mapping, ancestry inference, …) (?) • Gene therapy (?)

8. Successes and Lessons Learned • Attaching funding to community contributions can be effective • Tool developers need recognition (Bioinformatics application notes) • Comprehensive databases with simple, general visualization are catalysts for progress (Genome browsers) • Components and pipelines work • Progress has to be bottom-up as well as top-down

9. Current Obstacles • Incentives still not adequate in some cases: improving annotations, building well-engineered software, ensuring scientific reproducibility, obeying standards, etc. • Privacy and security are HUGE unresolved issues • Academic world still very focused on papers, but papers are static, machine nonreadable, incomprehensible to lay people • Hording of data for attribution or profit • Difficulty of interpreting results (personal genomics, phylogenetics, etc.)

10. Some Resources • NCBI gateway: www.ncbi.nlm.nih.gov • UCSC Genome Browser: genome.ucsc.edu • Ensembl Genome Browser: www.ensembl.org • PLoS Gateway: www.plos.org • BMC Gateway: www.biomedcentral.com • Personal genomics: www.23andme.com, www.navigenics.com, www.decodeme.com