Facilitating collaboration, exploration, and deployment for large-scale transcriptome data

# Facilitating collaboration, exploration, and deployment for large-scale transcriptome data
### Federico Marini (<a href="mailto:marinif@uni-mainz.de" class="email">marinif@uni-mainz.de</a>) University Medical Center Mainz
### 2021/10/28-29 Rhein-Main Cancer Retreat 2021 <a href="https://bit.ly/iSEE_RMCR"><code>https://bit.ly/iSEE_RMCR</code></a> <code> @FedeBioinfo</code>

---

layout: true

---

# About me

Hi, I'm **Federico**, and I like a lot to develop methods/software/apps (mainly in R/Bioconductor).

In my Bioinformatics group at the [IMBEI](https://www.unimedizin-mainz.de/imbei/) I try to combine all the above (and more) efficiently - often with a sprinkle of immunology and cancer biology.

<a href="mailto:marinif@uni-mainz.de">&nbsp; `marinif@uni-mainz.de`</a> 
<a href="https://federicomarini.github.io">&nbsp; `federicomarini.github.io`</a> 
<a href="http://twitter.com/FedeBioinfo">&nbsp; `@FedeBioinfo`</a> 
<a href="http://github.com/federicomarini">&nbsp; `@federicomarini`</a>

# About this talk

You can find this presentation here: 
[`https://federicomarini.github.io/RMCR2021/`](https://federicomarini.github.io/RMCR2021/) 
[`https://bit.ly/iSEE_RMCR`](https://bit.ly/iSEE_RMCR)

---

background-image: url("images/iSEE.png")
background-size: contain
background-position: 50% 50%

Available in Bioconductor [`https://bioconductor.org/packages/iSEE/`](https://bioconductor.org/packages/iSEE/)

---

# The analysis workflow

---

background-image: url("images/console_logcounts_sparse.png")
background-size: contain
background-position: 50% 50%

# The analysis workflow

Let's take a super cool project on single cell RNA-seq...

---

# The analysis workflow

<code>"Hi, I am a bioinformatician!" </code>

---

# The analysis workflow

<code>"And I am your next collaboration parner!" </code>

---

# The analysis workflow

<code>"You know what, I just had a cool idea about a project..." </code>

---

# The analysis workflow

---

# The analysis workflow

<code>This was indeed a productive kickoff meeting... </code>
---

# The analysis workflow

<code>"I've got a couple of more questions on the data..."</code>
---

# The analysis workflow

<code>"Well, it was more than a couple. But you can do analyses pretty quick right?" </code>
---

# The analysis workflow

<code>"Why don't you JUST do this tiny extra thing for me?" </code>

---

# The analysis workflow

<code>"... just do it yourself?" </code>

---

# Collaboration, exploration, deployment

**It takes (more than) two to tango** (at least, nowadays)

* You have results, but need to intepret them
* Your profiles are different: bioinformatician, life/medical scientist
* Everyone likes interactive things, but reproducibility of analysis is also key

You'll have...

better **quality assessment**

better **representation of the results**

better **generation of research hypotheses**

better **communication** of findings

higher **impact and reach** of your work

---

# Interoperable structures

---

# Interoperable structures

* `SummarizedExperiment`/`SingleCellExperiment`

* Seurat objects

* Scanpy + `AnnData`

With a few steps, they are (quite) interchangeable as well!

Excellent primer on single cell data analysis:  [`https://osca.bioconductor.org`](https://osca.bioconductor.org)!

???

iSEE focuses on the SingleCellExperiment class.
This class stores all the data and metadata associated with assays, cells, and features.

The SingleCellExperiment class is designed to accomodate all the information produced along a typical single-cell analysis workflow.

Those data include raw data:

- raw assay data
- experimental metadata
Processed data:
- quality control metrics
- normalized data
- dimensionality reduction results
Downstream analyses:
- cluster labels
- differential expression results
- downstream cell and feature annotations

---

# `iSEE` ❤️ SingleCellExperiment

Joint work with Aaron Lun, Charlotte Soneson, Kevin Rue-Albrecht initiated at [`#EuroBioc2017`](https://twitter.com/hashtag/EuroBioC2017)

"We could have an interactive SummarizedExperiment Explorer tool..."

---

# Using `iSEE` is easy

# `iSEE(se)`

*this is all you need, pinky swear*

---

background-image: url("images/tm_isee_allpanels.png")
background-size: contain
background-position: 50% 50%
class: center, bottom

# A tour of the panels

[`https://github.com/federicomarini/iSEE_instances`](https://github.com/federicomarini/iSEE_instances/tree/master/iSEE_tabulamuris)

---

background-image: url("images/gif_reordering.gif")
background-size: contain
background-position: 50% 50%
class: center, bottom

# Reordering the panels

---

background-image: url("images/tm_isee_reorderedpanels.png")
background-size: contain
background-position: 50% 25%
class: center, bottom

# Reordering the panels

---

background-image: url("images/gif_colorbygene.gif")
background-size: contain
background-position: 50% 50%
class: center, bottom

# Augmenting the observed data with biological knowledge

---

background-image: url("images/gif_linkedtotable.gif")
background-size: contain
background-position: 50% 50%
class: center, bottom

# Linking the panels

---

background-image: url("images/gif_customde.gif")
background-size: contain
background-position: 50% 50%
class: center, bottom

# Custom panels

---

background-image: url("images/gif_reprocode.gif")
background-size: contain
background-position: 50% 50%
class: center, bottom

# Code access for full reproducibility

---

background-image: url("images/gif_toursteps.gif")
background-size: contain
background-position: 50% 50%
class: center, bottom

# Interactive tours: an efficient means to communicate

---

# From zero to hero: `iSEE` the TICA atlas!

```r
library("iSEE")
library("zellkonverter")
sce_tica <- readH5AD("TICAtlas_downsampled_1000.h5ad")
HDF5Array::saveHDF5SummarizedExperiment(sce_tica, dir = "see_tica")
sce <- HDF5Array::loadHDF5SummarizedExperiment("see_tica/")
iSEE(sce)
```

---

# From zero to hero: `iSEE` the TICA atlas!

[`http://shiny.imbei.uni-mainz.de:3838/iSEE_TICA/`](http://shiny.imbei.uni-mainz.de:3838/iSEE_TICA/)

So, yes, `iSEE` also ❤️ `SeuratObject`s and `AnnData`'s h5ad format!

---

# What's new/coming for `iSEE` (>=2.6.0)?

- `iSEE(sce)` `\(\rightarrow\)` `iSEE()`  
  You can use `iSEE` as a local mini-server

- Additional biological layers
  * `iSEE` to work with other assay types (scATAC-seq specific functionality, e.g. via `episcanpy`)
  * multiomics datasets

- `iSEE` as a hub to serve many datasets as a portal (internal, public) - think of `cellxgene` and the SingleCellPortal, with much broader set of functionality (and lots of `iSEE_instances` as example)

- Got data? Let's `iSEE` that together...
  From a recent publication: [`http://shiny.imbei.uni-mainz.de:3838/iSEE_covidIT/`](http://shiny.imbei.uni-mainz.de:3838/iSEE_covidIT/)

- new API under the hood `\(\rightarrow\)` developing custom panels made easy - check out [`iSEEu`](https://bioconductor.org/packages/iSEEu/)

- the "Create next lab meeting presentation" button, to download all outputs at once

---

background-image: url("images/iSEE.png")
background-size: 150px
background-position: 90% 10%

# Summary

* **`iSEE`**: a simple, flexible, powerful, extendable framework 
 to **explore, deploy, work together** on data

* Bonus points: **reproducibility** of setup and code for scripted analyses + excellent didactical tool to generate such output

* Reduce the time-to-insight: **streamline** generation of findings and communication with collaborators.

* Enhance the medium where you can **tell a story with your data** - no need of static plots, welcome dashboards as a way to foster additional viewpoints

* Accompany your publications as live browser, promote **open data and open science**

* Voice control as a fun & accessible way to explore data

---

background-image: url("images/iSEE.png")
background-size: 150px
background-position: 90% 10%

# Links

* **[`https://bioconductor.org/packages/iSEE/`](https://bioconductor.org/packages/iSEE/)**

* **[`https://bioconductor.org/packages/iSEEu/`](https://bioconductor.org/packages/iSEEu/)**

* [`iSEE`](https://github.com/iSEE) organization
 + [`iSEE/iSEE`](https://github.com/iSEE/iSEE) source code
 + [`iSEE/iSEE2018`](https://github.com/iSEE/iSEE2018) manuscript
 + [`iSEE/iSEE-book`](https://github.com/iSEE/iSEE-book) repository for the book about extending `iSEE`
 + [`iSEE/iSEEu`](https://github.com/iSEE/iSEEu) source code for `iSEEu`, containing additional panels and modes for `iSEE`
 + [`iSEE/iSEE_custom`](https://github.com/iSEE/iSEE_custom) example of custom panels
 + [`iSEE/iSEE_instances`](https://github.com/iSEE/iSEE_instances) gallery of complete analyses on public data
 + [`iSEE/iSEEWorkshop2020`](https://github.com/iSEE/iSEEWorkshop2020) and [`iSEE/iSEEWorkshopEuroBioc2020`](https://github.com/iSEE/iSEEWorkshopEuroBioc2020), some extensive workshop materials - **Do you want us to "come to you" for a workshop?**

* Reinventing the wheel? [`https://github.com/federicomarini/awesome-expression-browser`](https://github.com/federicomarini/awesome-expression-browser)

---

## ... thank you for your attention!

Do you want to work with us? 
<code>marinif@uni-mainz.de - </code> [`@FedeBioinfo`](https://twitter.com/FedeBioinfo) 
<a href="https://federicomarini.github.io">&nbsp; `https://federicomarini.github.io`</a> 
<a href="http://www.imbei.de">&nbsp; IMBEI (Mainz, Germany)</a>

This presentation: [`https://federicomarini.github.io/RMCR2021/`](https://federicomarini.github.io/RMCR2021/) 
[`https://bit.ly/iSEE_RMCR`](https://bit.ly/iSEE_RMCR)

### Acknowledgements

IMBEI - Biostatistics & Bioinformatics division  
Center for Thrombosis and Hemostasis (CTH), Mainz (Virchow Fellowship)  
The `iSEE` development team & the Bioconductor Community for making this happen ❤️

---

---

## `iSEE(sce, voice = TRUE)`