G3 talk rld_2

Download G3 talk rld_2

Post on 10-Jul-2015

55 views

Category:

Science

0 download

Embed Size (px)

TRANSCRIPT

<ul><li><p>Tools for:Open-SourceOpen-Data</p><p>Rob L Davidson about.me/rob.davidson</p></li><li><p>The problem</p><p>reproducibility.cs.arizona.edu 515 papers (429 conf, 86 journal) </p></li><li><p>The problem</p><p>reproducibility.cs.arizona.edu</p></li><li><p>The Cause</p><p> Stodden 2010 638 registrant at NIPS</p><p> 30% share code 20% share data</p><p>http://web.stanford.edu/~vcs/papers/SMPRCS2010.pdf</p></li><li><p>Publishers must provide! HostingCurating</p><p>Citations for everything:data, tools + workflows</p></li><li><p>Tools for Reproducibility</p><p> Data: GigaDB Images: OMERO Workflows</p><p> Galaxy Executable Docs VMs </p></li><li><p>GigaDBgithub.com/gigascience/gigadb-cogini</p></li><li><p>Hosting all data</p></li><li><p>Hosting all research objects</p></li><li><p>Impact for research objects</p><p> Host Curate Share Cite - DOI</p></li><li><p>Even more accessible, transparent data?Hosting image data with OMERO</p></li><li><p>Hosting Images</p><p> Image LIMS Web embedding</p><p> View online, noneed for software</p><p> Full res Link all images to</p><p>publication No cherry picking</p><p>http://www.openmicroscopy.org/site/products/omero</p></li><li><p>NO</p><p>Cyber-Centipedes! Phenotyping</p></li><li><p>Accessible Cyber-Centipede images </p><p>OMERO: providingaccess to imaging data</p><p>View, filter, measure rawimages with direct linksfrom journal article.</p><p>See all image data, notjust cherry pickedexamples.</p><p>Download and reprocess.</p></li><li><p>OMERO: Adding value </p></li><li><p>The alternative... </p><p>...look but don't touch</p></li><li><p>Workflows 1. Galaxy</p><p>galaxyproject.org</p></li><li><p>galaxy.cbiit.cuhk.edu.hk</p></li><li><p>Implement workflows in a community-acceptedformat</p><p>http://galaxyproject.org</p><p>Over 45,000 main Galaxy server users</p><p>Over 1,000 papersciting Galaxy use</p><p>Over 55 Galaxyservers deployed</p><p>Open source</p></li><li><p>Copyright NBAF-B 2013Tool list Tool parameterisation Results panel</p><p>Implement workflows in an intuitive format</p></li><li><p>Visualising Workflows</p></li><li><p>Birmingham Metabo-Galaxy Workflow</p></li><li><p>Birmingham Metabo-Galaxy Tools wrapped in Python and XMLUser sees web form (easy!)Data stored centrally (secure!)Work done centrally (easy update)</p></li><li><p>Hosting Workflows</p></li><li><p>Hosting Workflows</p><p>1) Test data2) Software files3) Instructions+ Galaxy implementation</p></li><li><p>Can we reproduce results? SOAPdenovo2 S. aureus pipeline</p></li><li><p>Workflows2. Executable Docs</p></li><li><p>Open lab books, dynamic documents Facilitate reuse and sharing with tools like: Knitr, Sweave,</p><p>iPython Notebook</p><p>Sweave</p><p> Working towards executable papers</p></li><li><p>E.g.</p></li><li><p>E.g.</p></li><li><p>Some testimonials for KnitrAuthors (Wolfgang Huber)I do all my projects in Knitr. Having the textualexplanation, the associated code and the results all in oneplace really increases productivity, and helps explainingmy analyses to colleagues, or even just to my future self.</p><p>Reviewers (Christophe Pouzat) It took me a couple of hours to get the data, the fewcustom developed routines, the vignette and toREPRODUCE EXACTLY the analysis presented in themanuscript. With few more hours, I was able to modify theauthors code to change their Fig. 4. In addition to makingthe presented research trustworthy, the reproducibleresearch paradigm definitely makes the reviewers jobmuch more fun!</p></li><li><p>Workflow accessibility:VMs</p></li><li><p>Why VMs?</p><p> OS settings Dependencies</p><p> Versions e.g. python!</p><p> Data + Code linked Download or run in</p><p>cloud</p></li><li><p>VMs in GigaDB</p></li><li><p>Summary</p></li><li><p>Share data in GigaDBShare all images in GigaDB-View images via OMERO</p><p>Share code in GigaDB!Share pipeline using:</p><p>Executable docs!Galaxy!</p><p>VMs!</p></li><li><p>Give us data, papers&amp; pipelines*</p><p>Improvereproducibility!</p><p>scott@gigasciencejournal.com editorial@gigasciencejournal.com database@gigasciencejournal.com</p><p>Contact us:</p><p>* APCs currently generously coveredby BGI until 2015</p><p>www.gigasciencejournal.com</p></li><li><p>Ruibang Luo (BGI/HKU)Shaoguang Liang (BGI-SZ)Tin-Lap Lee (CUHK)Qiong Luo (HKUST)Senghong Wang (HKUST)Yan Zhou (HKUST)</p><p>Thanks to:</p><p>@gigasciencefacebook.com/GigaScienceblogs.biomedcentral.com/gigablog/</p><p>Peter LiHuayan Gao Chris HunterJesse Si ZheNicole NogoyLaurie GoodmanAmye Kenall(BMC)</p><p>Marco Roos (LUMC)Mark Thompson (LUMC)Jun Zhao (Lancaster)Susanna Sansone (Oxford)Philippe Rocca-Serra (Oxford) Alejandra Gonzalez-Beltran(Oxford) </p><p>www.gigadb.orggalaxy.cbiit.cuhk.edu.hk</p><p>www.gigasciencejournal.com</p><p>CBIITFunding from:</p><p>Our collaborators:team: Case study:</p><p>Slide 1Slide 2Slide 3Slide 4Slide 5Slide 6Slide 7Rewarding open dataSlide 9Slide 10Slide 11Slide 12Slide 13Slide 14Slide 15Slide 16Slide 17Slide 18Implement workflows in a community-accepted formatSlide 20Slide 21Slide 22Slide 23Slide 24Slide 25Slide 26Slide 27Slide 28Beyond Commenting Code: Step 2: Open lab books, dynamic documentsSlide 30Slide 31Some testimonials for KnitrSlide 33Slide 34Slide 35Slide 36Slide 37Slide 38Slide 39</p></li></ul>