Data Management for Quantitative Biology - Lecture 1, Apr 16, 2015
2 likes2,705 views
The document outlines the structure and requirements of a quantitative biology data management course at the University of Tübingen, detailing the course credits, grading criteria, and recommended literature. It covers key topics such as data management functions, technological innovations in biology, and methods in bioinformatics, emphasizing the importance of quantitative analysis in modern biological research. The lecture series includes various sessions on biological data management, database systems, and applications in quantitative biology.
Data Management for Quantitative Biology - Lecture 1, Apr 16, 2015
- 1. Dr. Sven Nahnsen, Quantitative Biology Center (QBiC) Data Management for Quantitative Biology Lecture 1: Introduction and overview
- 2. Overview • Administrative stuff (credits, requirements) • Motivation/quick review of relevant contents (Bioinformatics I and II) • Introduction to this lecture series • Semester overview 2
- 3. Course requirements To pass this course you must: • regularly and actively participate in the weekly problem sessions, • pass the final exam, assignments and projects • You have to work on assignments alone • You will work in small groups for the problem-orientated research project 3
- 4. Course credits and grading • Credits - MSc Bioinfo: 4 LP, module “Wahlpflichtbereich Bioinformatik” - MSc Info: 4 LP, area “Wahlpflichtbereich Informatik” • Grade - 30% assignments - 20% project - 50% finals • Finals: oral exam (30 minutes) covering the contents of the whole lecture, the assignments and the project • Finals will be scheduled at the end of the semester (Thu, 30/07/2015) 4
- 5. Recommended literature • We will point to relevant papers during the course of the literature • Important overview papers: § Hastings et al., 2005, Quantitative Bioscience for the 21st century. BioScience. Vol 55 No. 6 § Cohen JE (2004) Mathematics Is Biology's Next Microscope, Only Better; Biology Is Mathematics' Next Physics, Only Better. PLoS Biol 2(12): e439. • Books § Free E-Book: Data management in Bioinformatics ( http://en.wikibooks.org/wiki/Data_Management_in_Bioinformatics) § Lacroix, Z.; Critchlow, T. (eds): Bioinformatics: Managing Scientific Data. Morgan Kaufmann Publishers, San Francisco 2003 § Michale E. Wall, Quantitative Biology: From Molecular to Cellular Systems. 2012. Chapman & Hall § Pierre Bonnet. Enterprise Data Governance: Reference and Master Data Management Semantic Modeling. 2013. Wiley • Web resources § http://www.ariadne.ac.uk: Ariadne, Web Magazine for Information Professionals § http://www.dama.org: THE GLOBAL DATA MANAGEMENT COMMUNITY § H.D. Ehrich: http://www.ifis.cs.tu-bs.de/node/2855 5
- 6. Recommended Software • These software tools/framework and webservers will be used during the problem sessions http://www.cisd.ethz.ch/software/openBIS https://usegalaxy.org https://vaadin.com/home https://www.knime.org/ 6
- 7. Contact and organization • Questions concerning the lecture/assignments § dmqb-ss15@informatik.uni-tuebingen.de • Website § abi.inf.uni-tuebingen.de/Teaching/ws-2013-14/CPM • Christopher Mohr (Sand 14, C322) , Andreas Friedrich (Sand 14, C 304) • Dr. Sven Nahnsen (Quantitative Biology Center, Auf der Morgenstelle 10, C2P43, please send e-mail first) • Course material will be available on the website (see above), through social media channels and (if wished) as a hard copy during the lecture facebook.com/qbic.tuebingen twitter.com/qbic_tue 7
- 8. Who am I • Most of me and on our work can be found here: www.qbic.uni-tuebingen.de 8
- 9. Contents of this lecture Date Lecturer Lecture 8-‐10 AM Thursday 16 April 15 Nahnsen Introduc8on and overview Thursday 23 April 15 Nahnsen Biological Data Management Thursday 30 April 15 Czemmel Data sources ("Next-‐genera8on" technologies) Dr. Stefan Czemmel 9
- 10. Contents of this lecture Date Lecturer Lecture 8-‐10 AM Thursday 7 May 15 Codrea Database systems (mySQL, noSQL, etc.) Thursday 14 May 15 Ascension Day (Himmelfahrt) Thursday 21 May 15 Czemmel LIMS and E-‐Lab books Thursday 28, May 15 Kenar Experimental Design Dr. Marius CodreaErhan Kenar 10
- 11. Contents of this lecture Date Lecturer Lecture 8-‐10 AM Thursday 4 June 15 Corpus Chris8 Day (Fronleichnam) Thursday 11 June 15 Nahnsen Data analysis workflows (I) Thursday 18 June 15 Nahnsen Data analysis workflows (II) Thursday 25 June 15 Nahnsen Standardiza8on Thursday 2 July 15 Nahnsen Big Data Thursday 9 July 15 Nahnsen Integrated data management (OpenBIS, OpenBEB) Thursday 16 July 15 Nahnsen Applica8ons Thursday 23 July 15 Nahnsen Exam prepara8on Thursday 30 July 15 Nahnsen, Mohr, Friedrich EXAMS 11
- 12. What is your background? Ad hoc collection from the audience, Apr. 16, 2015 • Computer Science • Bioinformatics(immonoinformatics; User Front-end;integration , visualization) • Biology • Drug design • Agricultural biology (plant breeding) • Bioinformatics (Tx, NGS) • Geoecology • (ecology) • Biochemistry; Molecular Biology • Structural Biology • Electronic business 12
- 13. Let us brainstorm Ad hoc collection from the audience, Apr. 16, 2015 • What is data management? - Rapid access to data - Selective access to data; database queries - Combine data; manipulate data efficiently - Big data storage/analysis - Curating quality - Data visualization - Make data interpretable 13
- 14. Let us brainstorm • What is data management? http://zonese7en.com/wp-content/uploads/2014/04/Data-Management.jpg, accessed Apr 10, 2015, 11 AM 14
- 15. Data Management • The official definition provided by DAMA (Data management association) International, the professional organization for those in the data management profession, is: "Data Resource Management is the development and execution of architectures, policies, practices and procedures that properly manage the full data lifecycle needs of an enterprise.” • Further, the DAMA – Data management Body of Knowledge ((DAMA-DMBOK)) states:” Data management is the development, execution and supervision of plans, policies, programs and practices that control, protect, deliver and enhance the value of data and information assets ” Wikipedia: http://en.wikipedia.org/wiki/Data_management accessed Mar 30, 2015, 10 PM 15
- 16. 10 Data Management functions According to the DAMA Data Management Body of Knowledge (DMBOK) 16
- 17. Data governance • Strategy • Organization and roles • Policies and standards • Projects and services • Issues • Valuation Source: DAMA DMBOK Guide, p. 10 “Planning, supervision and control over data management and use” http://meship.com 17
- 18. Data quality management • Data cleansing • Data integrity • Data enrichment • Data quality • Data quality assurance Source: DAMA DMBOK Guide, p. 10 “defining, monitoring and improving data quality” http://www.arcplan.com/ 18
- 19. Data architecture management • Data architecture • Data analysis • Data design (modeling) Source: DAMA DMBOK Guide, p. 10 atasourceconsulting.com 19
- 20. Data development • Analysis • Data modeling • Database design • Implementation Source: DAMA DMBOK Guide, p. 10 dataone.org 20 “Data development is the process of building a data set for a specific purpose. The process includes identifying what data are required and how feasible it is to obtain the data. Data development includes developing or adopting data standards in consultation with stakeholders to ensure uniform data collection and reporting, and obtaining authoritative approval for the data set.”, A guide to data development, Australian Institute of Health and Welfare Canberra, 2007
- 21. Database management • Data maintenance • Data administration • Database management system Source: DAMA DMBOK Guide, p. 10 21
- 22. Data Security Management • Standards • Classification • Administration • Authentication/Authorization • Auditing Source: DAMA DMBOK Guide, p. 10 http://www.techieapps.com/wp-content/uploads/2013/07/2-1024x768.jpg 22
- 23. Reference and Master Data management • External/internal codes • Customer Data • Product Data • Dimension management (why do different dimensions (entities) relate to each other) • Taxonomy/Ontology Source: DAMA DMBOK Guide, p. 10 Master Reference 23 Reference data management
- 24. Master data
- 25. Data warehousing and business intelligence management • Architecture • Implementation • Training and Support • Monitoring and Tuning Source: DAMA DMBOK Guide, p. 10 Raw data Metadata … Summary data Data warehouse 25
- 26. Data warehousing and business intelligence management Raw data Metadata … Summary data Data warehouse Input Report Business intelligence 26
- 27. Document, record and content management • Acquisition and storage • Backup and Recovery • Content Management • Retrieval • Retention Source: DAMA DMBOK Guide, p. 10 27
- 28. Metadata management Metadata is data about data • Architecture • Integration • Control • Delivery Source: DAMA DMBOK Guide, p. 10 28
- 29. DAMA – DMBOK • A broad collection of all discipline and subtopics that are summarized under the umbrella of data management • These concern many business-related issues, but many concepts are very well applicable to the field of bioscience • We will come back to various aspects of the DAMA DMBOK during the course 29
- 30. Data management needs in science and research • Survey at the University of Oregon, USA (Brian Westra. "Data Services for the Sciences: A Needs Assessment". July 2010, Ariadne Issue 64 http://www.ariadne.ac.uk/issue64/westra/) • Different scientific discipline 30
- 31. Data management in science and research Brian Westra. "Data Services for the Sciences: A Needs Assessment". July 2010, Ariadne Issue 64 http://www.ariadne.ac.uk/issue64/westra/, accessed Apr. 10, 2015, 11 AM 1 2 3 4 5 6 7 8 9 10 11 1 Data storage and backup 7 Finding and accessing related data from others 2 Making scientific data findable by others 8 Connecting data storage to data analysis 3 Connecting data acquisition to data storage 9 Liniking this data to publications or other asset 4 Allowing or controlling access to scientific data by others 10 Ensuring data is secure and trustworthy 5 Documenting and tracking updates 11 Others 6 Data analysis and manipulation 31
- 32. Let us brainstorm • What is Quantitative Biology? Ad hoc collection from the audience, Apr. 16, 2015 - Not only yes/no, but put amounts to entities - Huge amount of data - Qunatitative methods to study biology - System-wide analysis; specific pathways - Make results human readable and accesible 32
- 33. Quantitative Biology • The term quantitative biology has been coined by Hastings et al., 2005. • High-throughput methods have led to a paradigm shift in biomedical research • Traditionally, the focus was on one-molecule-at-a-time for most bio(medical) research projects • Now, data on whole genomes, exomes, epigenomes, transcriptomes, proteomes and metabolomes can be generated at low cost. • The term quantitative biology is used to describe this paradigm shift. Improvements in this area have been driven mainly by two technological developments: Hastings et al., 2005, Quantitative Bioscience for the 21st century. BioScience. Vol 55 No. 6 33
- 34. Technological innovations • State-of-the-art mass spectrometers coupled to high- performance liquid chromatography through soft ionization techniques (HPLC-ESI-MS) have quickly changed the way we do proteomics, metabolomics, and lipidomics. • Next-generation sequencing has similarly changed the way we look at genomes, epigenomes, transcriptomes, and metagenomes. Due to advances in chemistry and imaging, sequencing reactions have been parallelized on a very large scale. The comprehensiveness of the data produced by high-throughput methods makes them particularly interesting as general-purpose analytical and diagnostic techniques. 34
- 35. Technological innovations • Imaging technologies can now produce high-resolution pictures of fine-grained cellular details at a very high speed • Finally methods from bioinformatics and computational biology have matured to rapidly analyze the huge raw data sets that are generated by these high throughput technologies 35
- 36. Contents from Bioinformatics 2 (high-throughput technologies • Most of the high throughput technologies have been introduced during the Bioinformatics II lecture • There are specialized lectures on “Transcriptomics” and on “Computational Proteomics and Metabolomics” • We will give a short Recap on the Bioinformatics II contents that are relevant for this lecture • More advanced topics on data generation methods will be introduced in lecture 3 by Dr. Stefan Czemmel (focus on next generation sequencing) 36
- 37. Origin of the “Central Dogma of Molecular Biology” (Francis Crick, 1956) The central dogma of molecular biology • First articulation by Francis Crick in 1956 • Published in Nature in 1970 37
- 38. The central dogma – classical view • In general, the classic view reflects how biology is (biological data are) organized • Genomics, however, enabled a more complex view Cox Systems Biology Lab | Research, University of Toronto, Canada 38
- 39. Reminder (Bioinformatics 2) Oltvai-Barabasi, Science, 2002 39
- 40. Recap Bioinformatics II: Systems biology • Quantitative data on various levels of biological complexity build fundaments of systems biology • Mathematical modeling has been based on gene expression • Recent important technological improvements allow the analysis of protein and metabolite profiles to a great depth • Important layers for understanding biology • New experimental techniques offer tremendous challenges for computational analysis 40
- 41. Recap Bioinformatics II: Aims of Systems Biology • Describe large-scale organization • Quantitative modeling • Describe cell as system of networks - Fundamental research: time-resolved quantitative understanding of living systems - Medicine: enable personalized medicine (e.g., improve treatment strategies for cancer patients) - Biotechnology: improve production, degradation, construction of synthetic organisms, etc. 41
- 42. Exp. Methods – Transcriptomics • Extract and amplify RNA • Hybridization on microarray • Identify and quantify by fluorescence signal • Sequences can be mapped back to genome Lindsay, Nature Rev. Drug Discovery, 2003, 2, 803 42
- 43. Microarray Data Analysis • Key problems in microarray data analysis are - Data normalization - Clustering - Dimension reduction - Diagnostics/classification - Network inference - Visualization of results Janko Dietzsch , Nils Gehlenborg and Kay Nieselt. Mayday-a microarray data analysis workbench. Bioinformatics 2006 22(8):1010-1012 43
- 44. Genome sequencing February 15, 2001 February 16, 2001 44
- 45. Genome sequencing • 2001: initial publication • 2003: 2nd draft “Human Genome” • > 13 years of work and > 3*109 $ • 2010: 8 days 1*104 $ • Today: approximately 5.5 days and < 1*104 $ • Future: within 3 years Biotech company (Pacific Biosciences) expects similar amount of data in < 15 min for < 1*103 $ 45
- 46. Status genomics/transcriptomics • Dramatic drop in cost for genome sequencing • Number of sequenced genomes grows continuously • Genome is a very static snapshot of living system • Biological adaption is rather slow; long-term information storage • Proteins and their reaction products, metabolites are much closer to reality • Genome and transcriptome databases are essential bases for proteomics and metabolomics research 46
- 47. Genomics vs. Proteomics Genomics Proteomics Genomes rather static ~ 20 k genes established technology (capillary sequencer) Proteomes are dynamic (age, tissue, breakfast, …) up to 1000 k proteins emerging technologies (MS, HPLC/MS, protein chips) 47
- 48. Proteomics http://www.iamashcash.com/wp-content/uploads/2011/03/caterpillar-to-butterfly1.jpg, accessed: 14/10/2013 6 PM Genome remains the same Proteome changes 48
- 49. Main fields of proteomics 49
- 52. Next generation sequencing 1st generation 2nd generation Illumina / Solexa Genetic Analyzer 2000 Mb / run (96h) Roche / 454 Genome Sequencer FLX 400 Mb / run (8h) Applied Biosystems SOLiD 3000 Mb / run (120h) 300 : 1 (2008) Applied Biosystems 3730xl 0,08 Mb / run 1 Mb / 24 h >3000 : 1 (2010) 1st generation 2nd generation Slide: Prof. Peter Bauer In 2008 52
- 53. „3rd“ generation sequencing Drmanac Science (2010) 5961: 78-81. CompleteGenomics DNB sequencing 18x 210Gb / run >37.000 : 1 (2010) 3rd generation sequencing Slide: Prof. Peter Bauer 53
- 54. High resolution imaging “Imaging in biology may refer to >15 different technologies” Prominent and data-intense examples include: • Optical (bioluminescence and fluorescence imaging) • Magnetic resonance imaging • X-ray computed tomography • Positron emission tomography http://en.wikipedia.org/wiki/Biological_imaging, accessed Apr. 13, 4 PM 54
- 55. Imaging workflow Eliceiri et al., Nature Methods 9, 697–710 (2012) 55
- 56. Database systems • Relational databases - Example MySQL • NoSQL databases - Example MongoDB • How to query databases • Entity relationship models • Repositories (e.g. Pride, PeptideAtlas) - Annotations - Sequences 56
- 59. Large-scale study data – 1000 Genomes • Sample lists and sequencing progress • Variant Calls • Alignments • Raw sequence files http://www.1000genomes.org/data 59
- 60. Large-scale study data – The cancer genome atlas (TCGA) • TCGA aims to help to diagnose, treat and prevent cancer • explore the entire spectrum of genomic changes involved in more than 20 types of human cancer. • Approx. 2 PB of genomic raw data http://cancergenome.nih.gov 60
- 61. Laboratory information management systems/ Electronic Lab Books • How to track all information that is generated in the laboratory • Automated annotation of all experimental parameters is essential for reproducible science • Currently, most experiments are protocolled manually in lab textbooks • Data security (intellectual property versus open data) 61
- 62. Experimental design • Biological experiments are very complex • Statistical significance requires a high number of biological replicates • Often many different conditions and time points need to be considered • One study can involve many different experiments (multi omics studies involve different omics layers, e.g. genomics + transcriptomics + proteomics) • All experiments come with different meta data requirements • For various reasons the experimental design is not always balanced (e.g. 5 samples in group A and and only 3 samples are available for group B) Friedrich, A., et al. Biomed Research International, April 2015 – in press. Nahnsen, S., Drug Target, May 2015 – in press. 62
- 63. Experimental design Friedrich, A., et al. Biomed Research International, April 2015 – in press. Nahnsen, S., Drug Target, May 2015 – in press. 63
- 64. Data analysis workflows • Chain different (heterogeneous) tools • Parameter handling • Execution in high performance computing environment made easy 64
- 65. Standardization in bioinformatics • Many world-wide bioinformatics initiatives need to rely on open standards • Development of standards has to be a community effort • Standardized data formats are important to guarantee - Sustainability - Independence of instrument vendors - Independence of analysis software - Exchangeability of raw data • Standard formats increase the amount of data by a factor of x (x = 2-4) • Many people refrain from using open standards 65
- 66. http://en.wikipedia.org/wiki/Big_Data, accessed Apr 24, 2014 Big data is the term for a collection of data sets so large and complex that it becomes difficult to process using on-hand database management tools or traditional data processing applications. The challenges include capture, curation, storage, search, sharing, transfer, analysis and visualization. The trend to larger data sets is due to the additional information derivable from analysis of a single large set of related data, as compared to separate smaller sets with the same total amount of data, allowing correlations to be found …… Big data 66
- 67. Big data examples • European Council for Nuclear Research (CERN) Geneva, Swizerland 25 Petabyte/Jahr at LHC (Large Hadron collider) (~6.2 Mio. DVDs) CERN LHD data Big data Beispiele ep.ph.bham.ac.uk, 2014 67
- 68. Big data examples • Google verarbeitet 9.1 Exabyte/year (300 Mio. DVDs) GOOGLE data Mayer-Schönberger, 2013, ititch.com, 2014 68
- 69. Biology and Big data? • Klassisch: Beobachtung der Natur und deren Phänomene DNA RNA Proteine Träger der Erbinformation Expression von bestimmten Genen Üben nötige Funktion in der Zelle aus 1956 formuliert Francis Crick das zentrale Dogma der Molekularbiologie: • 1950er JahreDurchbruch in der Molekularbiologie 69
- 70. Big data Vivien Marx, Biology: The big challenges of big data, Nature. 2013, doi:10.1038/498255a 70
- 71. Integrated data management in biology/biomedicine 71 http://media.americanlaboratory.com/m/20/Article/35231-fig1.jpg
- 73. NGS Lab Lab Storage Data movers • Automatically moves large to huge file-based data to a remote (central) storage • Uses rsync routine; easy configuration using config file • Data mover athentification: public/private key ssh authentification • Moves data to openbis dropboxes (individual boxes and users for each of the five member labs) Data Mover DataMover: • Developed at ETH Zurich as part of OpenBIS • http://www.cisd.ethz.ch/software/Data_Mover 73
- 74. openBIS (meta) data store • Open, distributed system for managing biological information • Captures different experiment types (OMICS, imaging, screening,...) • Tracking, annotating and sharing of experiments, samples and datasets for distributed research • Different servers for meta data and bulk raw data • Underlying PostgreSQL database • ETL routines for extraction of meta data and linking 74
- 77. Applications • Personalized medicine: Individualized vaccination in cancer • Large-scale clinical studies: example Hepatocellular carcinoma 77
- 78. Contact: Quantitative Biology Center (QBiC) Auf der Morgenstelle 10 72076 Tübingen · Germany dmqb-ss15@informatik.uni-tuebingen.de Thanks for listening – See you next week