Skip to content

Latest commit

 

History

History
565 lines (429 loc) · 16.9 KB

File metadata and controls

565 lines (429 loc) · 16.9 KB
title theme revealOptions css
OMERO and the Image Data Resource: Supporting open science in Dundee and around the world
white
transition center slideNumber showSlideNumber
fade
false
c/t
speaker
css/ome-reveal.css

OMERO and the
Image Data Resource:

Supporting open science in Dundee and around the world

Simon Li
Open Microscopy Environment

Open science

Who thinks it's a good idea?

More than 70% of researchers have tried and failed to reproduce another scientist's experiments, and more than half have failed to reproduce their own experiments.

http://www.nature.com/news/1-500-scientists-lift-the-lid-on-reproducibility-1.19970


Why is the IDR needed?

Bioinformatics databases are taken for granted in the life sciences

![ensemble](images/ensembl-homepage.png) ![pdf](images/pdb-homepage.png) ![ncbi](images/ncbi-homepage.png)

What about imaging?

Image databases have lagged behind though. There are several application-specific repositories:

![emdb](images/emdb-homepage.png) ![lincs](images/lincs-homepage.png) ![jcbdataviewer](images/jcbdataviewer-homepage.png)

What's the difficulty?

cell


What's the difficulty?

>sp|Q96GD4|AURKB_HUMAN Aurora kinase B OS=Homo sapiens GN=AURKB PE=1 SV=3
MAQKENSYPWPYGRQTAPSGLSTLPQRVLRKEPVTPSALV
LMSRSNVQPTAAPGQKVMENSSGTPDILTRHFTIDDFEIG
RPLGKGKFGNVYLAREKKSHFIVALKVLFKSQIEKEGVEH
QLRREIEIQAHLHHPNILRLYNYFYDRRRIYLILEYAPRG
ELYKELQKSCTFDEQRTATIMEELADALMYCHGKKVIHRD
IKPENLLLGLKGELKIADFGWSVHAPSLRRKTMCGTLDYL
PPEMIEGRMHNEKVDLWCIGVLCYELLVGNPPFESASHNE
TYRRIVKVDLKFPASVPMGAQDLISKLLRHNPSERLPLAQ
VSAHPWVRANSRRVLPPSALQSVA

What's the difficulty?

  • Image data is extremely difficult to automatically annotate
    • There are 100s of different file formats
  • Genomics/proteomics data is often text-based
    • annotations are mandatory for understanding

What is the Image Data Resource?



A repository for reference datasets and images that

“constitute valuable resources for a broader community of users that will often be accessed as a reference or that will be recomputed to extract additional information and knowledge”

Euro-BioImaging - ELIXIR Image Data Strategy


The IDR today:

  • 30 data submissions
  • 1 million experiments
  • 2.7 million images
  • 37 million image planes
  • 43 TB raw data
  • 15 million files
  • All data is carefully curated


Previously inaccessible datasets

![idr0001-example](images/big_data_harddrives.jpg)]

idr0001-example


Genetic HCS

idr0008a-example

idr0008a-example


Geographic HCS

![idr0015-example](images/idr-datasets/idr-0015-image-1976668.jpg) ![idr0015-example](images/idr-datasets/idr0015-example-external2.png)

![idr0008a-example](images/idr-datasets/idr-0008a-plate-230.png) Genetic HCS
![idr0015-example](images/idr-datasets/idr-0015-plate-4871.png) Geographic HCS
![idr0006-example](images/idr-datasets/idr-0006-plate-2243.png) Gene product targeting HCS
![idr0017-example](images/idr-datasets/idr-0017-plate-4263.png) Chemical HCS
![idr0018-example](images/idr-datasets/idr-0018-image-1919122.jpg) Histopathology
![idr0021-example](images/idr-datasets/idr-0021-image-1885136.jpg) 3D‑SIM
![idr0023-example](images/idr-datasets/idr-0023-image-1886132.jpg) Super-resolution
![idr0027-example](images/idr-datasets/idr-0027-image-2858285.jpg) Chromatin dynamics

Genes and phenotypes in the IDR

Image 2971124


<iframe src="https://idr.openmicroscopy.org/mapr/gene/?experimenter=-1" class="stretch"></iframe>


How do we do this?

![Eleanor Williams](images/team/eleanor_williams.png)
  • 609 MB of spreadsheets
  • 334 MB of text files
![idr0020-study.txt](images/spreadsheets/idr0020-study.png) ![idr0020-screenA-annotation.csv](images/spreadsheets/idr0020-annotation.png) ![idr0020-screenA-library.txt](images/spreadsheets/idr0020-library.png)

What about your datasets?


Hi-res live cell imaging of chromatin dynamics

idr0027-paper


UOD Life Science's own OMERO server

  • Open to everyone in SLS (and others on request)
  • You can choose to publish some of your images (viewable without a login)

Schleicher 2017

Katharina Schleicher


OMERO.figure


<iframe data-src="https://omero.lifesci.dundee.ac.uk/figure/file/364215/" class="stretch"></iframe>

So far we have ...

  • A public curated reference image repository with internal and external cross-references
  • An OMERO server for publishing datasets within Dundee
  • OMERO.figure, a full pipeline for creating publication quality figures

What's next?

Tools for analysing image data in the IDR and OMERO

  • Reproducibly
  • And shareably
  • But why only in OMERO? And why only images?

Analysis in the cloud

An online cloud-based analysis platform for running and sharing analysis scripts

Python R



Balaji

Example: searching the IDR for novel gene interactors

  1. Start with a set of genes we're interested in
  2. Obtain a phenotypic profile characterising this set of genes
  3. Query the IDR for other genes with a similar phenotypic profile
  4. Hypothesis: these genes are potentially associated with the complex

1. Start with a set of genes we're interested in

  • Arp2/3 protein complex Jupyter ARP2-3 variables

2. Obtain a phenotypic profile characterising this set of genes

  • Search all studies in the IDR, obtain all phenotypes associated with these genes Jupyter ARP2-3 query phenotypes

3. Query the IDR for other genes with a similar phenotypic profile

  • This gives us new targets to investigate
![Jupyter ARP2-3 similar genes](images/jupyter/idr-arp23-similargenes.png)

4. Hypothesis: these genes are potentially associated with the complex

  • Initial validation with StringDB Jupyter ARP2-3 interactors code

Remember: this is very
preliminary work!


Anyone can have access to:

![idr](images/idr-about.png) https://idr.openmicroscopy.org
![nightshade](images/summary/nightshade-web-1.png) https://nightshade.openmicroscopy.org
![figure](images/summary/nightshade-figure-example.png) OMERO.figure
![analysis](images/summary/idr-jupyter-example.png) IDR cloud analysis

Contact us for more details: https://www.openmicroscopy.org/support/


![Jason Swedlow](images/team/jason_swedlow.jpg) Jason Swedlow
![Josh Moore](images/team/josh_moore.jpg) Josh Moore
![Simon Li](images/team/simon_li.png) Simon Li
![Eleanor Williams](images/team/eleanor_williams.png) Eleanor Williams
![Gabriella Rustici](images/team/gabriella_rustici.png) Gabriella Rustici
![David Gault](images/team/david_gault.png) David Gault
![Dominik Lindner](images/team/dominik_lindner.png) Dominik Lindner
![Harald Waxenegger](images/team/harald_waxenegger.png) Harald Waxenegger
![Helen Flynn](images/team/helen_flynn.png) Helen Flynn
![Jean-Marie Burel](images/team/jeanmarie_burel.jpg) Jean-Marie Burel
![June Matthew](images/team/june_matthew.png) June Matthew
![Kenny Gillen](images/team/kenny_gillen.jpg) Kenny Gillen
![Mark Carroll](images/team/mark_carroll.jpg) Mark Carroll
![Petr Walczysko](images/team/petr_walczysko.png) Petr Walczysko
![Roger Leigh](images/team/roger_leigh.jpg) Roger Leigh
![Sebastien Besson](images/team/sebastien_besson.jpg) Sebastien Besson
![Will Moore](images/team/will_moore.png) Will Moore
![OME](images/team/logo-ome.svg) ![University of Dundee](images/team/logo-dundee.png)
![Rafael Carazo-salas](images/team/rafael_carazosalas.png) Rafael Carazo-salas
![University of Bristol](images/team/logo-bristol.png)
![BBSRC](images/logos/bbsrc.png)
![Wellcome Trust](images/logos/wellcome.png)
![Alvis Brazma](images/team/alvis_brazma.png) Alvis Brazma
![Ugis Sarkans](images/team/ugis_sarkans.png) Ugis Sarkans
![Simon Jupp](images/team/simon_jupp.png) Simon Jupp
![Tony Burdett](images/team/tony_burdett.png) Tony Burdett
![ebi](images/team/logo-ebi.png)
![Aleksandra Tarkowska](images/team/aleksandra_tarkowska.jpg) Aleksandra Tarkowska
![Richard Ferguson](images/team/richard_ferguson.png) Richard Ferguson
![Simone Leo](images/team/simone_leo.jpg) Simone Leo
![Bálint Antal](images/team/balint_antal.png) Bálint Antal
![Anatole Chessel](images/team/anatole_chessel.jpg) Anatole Chessel