Advertisement
Advertisement

More Related Content

Advertisement

More from ICRISAT(20)

Advertisement

ICRISAT Global Planning Meeting 2019: Research Data Management by Abhishek Rathore

  1. Data Management 4 7th Feb 2019 ICRISAT Global Planning Meeting ICRISAT Patancheru, India Abhishek Rathore Principal Scientist & Theme Leader Statistics, Bioinformatics & Data Management a.rathore@cgiar.org Data Driven Decisions & Big Data
  2. Work with researchers to achieve higher “research”, 1) Quality 2) Credibility / Repeatability 3) Cost-effectiveness by Leveraging Modern Data Science SBDM Vision
  3. Research Data Management 8. Data Querying, Archiving & Sharing Conclusion of Proposal 7. Valid Interpretation & Publication of Results Execution of Proposal 6. Statistical Modelling & Data Analysis Execution of Proposal 5. Data QC / Curation Process & Record Execution of Proposal 4. Data Generated / Captured & Stored Centrally Execution of Proposal 3. Setup Experiment/ Study/ Survey/ Data Protocols/ Timelines/ Ontologies/ Sample Tracking/ Barcodes/ ODK/ Electronic Books Execution of Proposal 2. Collection/Generation/Infrastructure/Timelines Submission of Proposal 1. Project Hypothesis & Data Requirement Submission of Proposal
  4. Data Management Infrastructure Phenotypic – Trials • BMS of IBP • Amazon Cloud - EC2 • Support of APIs Data Points: 1,200,000 Genotypic Data • GOBii • Installed on Cluster • Support of APIs Data Points: 400 Billion High-Throughput Phenotyping Data Points: >50 Million/Year • Online Server • No APIs
  5. Data Points: >91 Million • No ontology based system • Custom servers • Difficult cross query • No APIs Data Management Infrastructure Socio-Economic Data Genebank Data • Custom solutions / servers • Difficult cross query • No APIs Weather Data • Interactive Dashboard • Difficult to cross query • No APIs
  6. Data Management Infrastructure Soil/Agronomy/GIS/On-farm • Soil Samples – Dataverse Files / R Accessible • DNA Samples – BMS/GOBii • Grain Quality – BMS • Plant Sample – BMS • On-Farm – Dataverse Files / R Accessible • And: Many other data base platforms & data sets Data Points: > 10,000,000 ICRISAT well positioned in handling BigData
  7. Data Management Infrastructure More on BMS • Seamless Data Transfer • Extremely Simple – 5 Clicks • Intermediate Save Possible • Several ways of data upload • Data Clinic Tomorrow • Email a.rathore@cgiar.org
  8. Data Management Infrastructure More on Digitalization/Automation Packet Printer Seed Counters with Digital Balance Harvest Masters
  9. Drones, IoT – Data & Analytics Pipeline • Drone Data Management & Analytical Pipeline • Pilot with CIAT and Cornell • SBDM initiated EiB & BigData Platform • Worked on breeding use cases • Workshop in Tanzania • Drones Working Group EiB M5 • CoP on Biometrics & Bioinformatics IoT Platform Standardization • Temperature • RH • Day Light • Moisture
  10. More on Data Management BMS & Open Data Workshops 1 ICAR-ICRISAT collaborative workshop to strengthen breeding modernization activities 25-27 July2018 Lilongwe Malawi 2 Data Hackathon for ESA and WCA Regions 23-24 July2018 Lilongwe Malawi 3 Breeding Data Management for HOPE II and TL III NARS for ESA 16-20 July2018 Debrezeyit Ethiopia 4 Workshop on modernization of breeding programs through BMS at ICAR-Indian Institute of Pulses Research (IIPR), Kanpur, India during 02-06 July, 2018 02-06 July2018 ICAR-IIPR, Kanpur India 5 BMS and Data Archiving training workshop for TLIII and HOPE II partners 18-22 June 2018 ICRISAT- Samanko, Bamako Mali 6 Data Hackathon for ASIA 15-16 March 2018 ICRISAT- Patancheru India Data Analytics Workshops (EiB, BigData & GLDC) 1 2nd International Workshop on Advanced R & R/QTL 03-07 December 2018 ICRISAT- Patancheru India 2 EiB Module 5 B&B CoP Annual Collaborative Workshop 24-29 September 2018 Amsterdam The Netherlands
  11. More on Data Management CGIAR Big Data Annual Convention 2019 16th-18th October 2019 Venue : TBD, Hyderabad ICRISAT is hosting
  12. Data Management Infrastructure Open Access Data • ICRISAT Technologically Modern • Data Platforms are Modern • However, Platforms are Nothing without DATA! • ICRISAT investing in Open Data • Dataverse • Hackathons Organized • Educating Researchers • Open Data Need • Metadata standards • Curation • Best Practices
  13. How Big is BIG?
  14. How Big is BIG? BIG data problem @ ICRISAT?
  15. Big Data Management Infrastructure
  16. Big Data Management Infrastructure
  17. Big Data Platforms Use Case - GS ICRISAT Only Center
  18. Data Management Infrastructure Future Plans • Centralize Data Management System Across ICRISAT • Reduce Micro Databases - APIs • Develop APIs and Wrappers • Virtualize Platforms - Scalability • Develop Dashboards • Custom Interfaces • Define Big Data Analytics Use cases • Researchers Inputs • Stakeholders Inputs • And anything what you all want!
  19. Data Management Infrastructure Future Plans
  20. Data Quiz 1 According to Approved Data Management Policy are we supposed to submit a Data Management Plan, Budget & Discuss with SBDM?
  21. Data Quiz 2 According to Approved Data Management Policy after how many months (MAX) researchers are supposed to make their data open & submit to SBDM?
  22. Data Quiz 3 According to Approved Data Management Policy Who owns data? Researchers or Institute? And should it be made clear in contract?
  23. Digital Strategy - ICRISAT Data Management Step Project Hypothesis & Data Requirement: Define hypothesis & decide what data is needed to test hypothesis Risk (if not done properly) Data may not be sufficient to test hypothesis Present Status Only 2%-3% project discuss data management plan Future Status Each project must discuss data and hypothesis and must have a well discussed data management plan in place at time of submission of project
  24. AhsanteSana… Dhanyavad… Merci… NaGode Salamat… Shukran… Thanks…
Advertisement