Shiny Apps in Genomics and Trials

class: center, middle, inverse, title-slide

# Shiny Apps in Genomics and Trials
## R/Pharma 2018
<div style='float:left'>
<hr color='#EB811B' size=1px width=720px>
</div>
 
### Jessica Minnier, PhD Oregon Health & Science University @datapointier
### August 16, 2018 Slides available at <a href="http://bit.ly/rpharma-minnier" class="uri">http://bit.ly/rpharma-minnier</a>

---

# Our Team

## Biostatistics Shared Resoure

<!---

# Some Examples

## START: Shiny Transcriptome Analysis Resource Tool

RNA-seq/Microarray analysis and visualization

## Drug Combination in Cancer

Drug development

## Adaptive Bayesian Design

Clinical Trials Design
--->
---

class: inverse, center, middle

# Why Shiny?

---

background-image: url(figs/spreadsheetcombo.png)
background-position: center
background-size: 120%

# Motivation: Too many spreadsheets!

---

# An example: START

Shiny 
Transcriptome
Analysis
Resource
Tool

BSR: Jessica Minnier, PhD;
OHSU KCVI: J Sklenar, PhD, A.P. Barnes, PhD, J. Nelson, PhD
 
Github: https://github.com/jminnier/STARTapp/
 
Shinyapps.io: https://kcvi.shinyapps.io/START/
 

DATA = RNA-seq gene expression

---

<!---

# Very quick tour

--->

# Very quick tour

---

# Data Input

- Input data ~ multiple formats: Special input formats
- Interface with other software? Export special formats

---

# Data Exploration: Interactive

![](figs/start_scatterplot.png)
---

# Drug Development

## Combination Therapy `\(\rightarrow\)` Visualize drug synergy

*BSR*: Andy Kaempf, MS & Tomi Mori, PhD 
*Knight Cancer Leukemia Team*: Steve Kurtz, PhD & Chris Eide, B.A.

![](figs/combo1.png)

---
class: clear

![](figs/combo4.png)

---

# Clinical Trials Methods

## Bayesian Adaptive Design

`\(\rightarrow\)` An Adaptive Bayesian Sequential Design for Determining the Subject-Specific Optimal Treatment Length
(manuscript in progress)

https://bayesiantrial.shinyapps.io/Bayesian12/

*BSR*: Yiyi Chen, PhD; Lina Gao, PhD 
*OHSU*: Zunqiu Chen, MS; Matthew F Halsey, PhD; Jodi Lapidus, PhD

- Select directions and weights of various outcomes
- Determine potential length of treatment to be evaluated
- Upload data
- Potential actions: Escalate; De-escalate; Stay

![](figs/bayes0.png)
---

# General App & Trial Specific

![](figs/bayes1.png)

---
class: inverse, middle, center
# Tips/Lessons Learned

---

# Present QA plots up front

---

# Provide Instructions!

---

# Old fasioned debugging/logging works

![](figs/start_error.png)

---

# Conclusions

- R Shiny can be a tool to creatively interact with data
- Improve collaboration and promote your methods
- Avoid repetitive analyses and visualizations
- Instructions, data checks/QA are vital

## START

- working on extensions to single-cell and proteomics specific analyses and visualizations
- (one day) switch to use modules
- we welcome collaboration!

---

# Thank you!

Contact me: minnier-[at]-ohsu.edu, [datapointier](https://twitter.com/datapointier), [jminnier](https://github.com/jminnier/)

START: http://jessicaminnier.com/STARTapp/

RELATED SLIDES ON SHINY (Challenges/Responsibilities):
- Data Day Texas <http://bit.ly/shiny-ddtx>
- CSP: <http://bit.ly/shiny-csp18>

Slides available at <http://bit.ly/rpharma-minnier>

Code for slides available at <https://github.com/jminnier/talks_etc>

Slides created via the R package [xaringan](https://github.com/yihui/xaringan) by [Yihui Xie](https://twitter.com/xieyihui?lang=en) with the metropolis theme